Narrative and Identity in the Ancient Greek Novel: Returning Romance (Greek Culture in the Roman World)

Citation preview

This page intentionally left blank

NARRATIVE AND IDENTITY IN THE ANCIENT GREEK NOVEL

The Greek romance was for the Roman period what epic was for the archaic period or drama for the classical: the central literary vehicle for articulating ideas about the relationship between self and community. This book offers a fresh reading of the romance both as a distinctive narrative form (using a range of narrative theories) and as a paradigmatic expression of identity (social, sexual and cultural). At the same time, it also emphasises the elasticity of romance narrative, its ability to accommodate both conservative and transformative models of identity. This elasticity manifests itself partly in the variation in practice between different romancers, some of whom are traditionally Hellenocentric and others more challenging; but ultimately, it is argued, it reflects a tension in all romance narrative, which characteristically balances centrifugal against centripetal dynamics. This book will interest classicists, historians of the novel, and students of narrative theory. t i m wh i t m a r s h is a leading literary and cultural critic of the Greek world during the time of the Roman empire. A specialist in both ancient texts and modern theories, he has written over fifty articles and five books, including Greek Literature and the Roman Empire (2001) and The Second Sophistic (2005); he has also edited or co-edited four books, and edits two book series.

G REEK CULTURE IN THE RO M A N WO RL D Editors susan e. alcock, Brown University ja S´ elsner, Corpus Christi College, Oxford simon goldhill, University of Cambridge The Greek culture of the Roman empire offers a rich field of study. Extraordinary insights can be gained into processes of multicultural contact and exchange, political and ideological conflict, and the creativity of a changing, polyglot empire. During this period, many fundamental elements of Western society were being set in place: from the rise of Christianity, to an influential system of education, to long-lived artistic canons. This series is the first to focus on the response of Greek culture to its Roman imperial setting as a significant phenomenon in its own right. To this end, it will publish original and innovative research in the art, archaeology, epigraphy, history, philosophy, religion, and literature of the empire, with an emphasis on Greek material.

Titles in series: Athletics and Literature in the Roman Empire Jason K¨onig Describing Greece: Landscape and Literature in the Periegesis of Pausanias William Hutton Religious Identity in Late Antiquity: Greeks, Jews and Christians in Antioch Isabella Sandwell Hellenism in Byzantium: The Transformations of Greek Identity and the Reception of the Classical Tradition Anthony Kaldellis The Making of Roman India Grant Parker Philostratus Edited by Ewen Bowie and Ja´s Elsner The Politics of Munificence in the Roman Empire: Citizens, Elites and Benefactors in Asia Minor Arjan Zuiderhoek Saints and Church Spaces in the Late Antique Mediterranean: Architecture, Cult, and Community Ann Marie Yasin Galen and the World of Knowledge Edited by Christopher Gill, Tim Whitmarsh and John Wilkins Local Knowledge and Microidentities in the Imperial Greek World Edited by Tim Whitmarsh Homer Between History and Fiction in Imperial Greek Literature Laurence Kim Epiphany and Representation in Graeco-Roman Culture: Art, Literature, Religion Verity Platt Narrative and Identity in the Ancient Greek Novel Tim Whitmarsh

NARRATIVE AND IDENTITY IN THE ANCIENT GREEK NOVEL Returning Romance

by TIM WHITMARSH

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, S˜ao Paulo, Delhi, Tokyo, Mexico City Cambridge University Press The Edinburgh Building, Cambridge cb2 8ru, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521823913  c Tim Whitmarsh 2011

This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2011 Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Whitmarsh, Tim. Narrative and identity in the ancient Greek novel : returning romance / Tim Whitmarsh. p. cm. – (Greek culture in the Roman world) Includes bibliographical references and index. isbn 978-0-521-82391-3 (hardback) 1. Greek fiction – History and criticism. 2. Narration (Rhetoric) – History – To 1500. I. Title. pa3267.w55 2011 883 .0109 – dc22 2010052774 isbn 978-0-521-82391-3 Hardback

Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

½xe±v o¬ tän –rÛntwn logismoª kaª pr»ceiroi ËpopteÓsai, deinoª d• e«k†sai, ›nqeoi d• manteÅsasqai. Keen are the ratiocinations of lovers, and quick to suspect; clever at guessing, and inspired at predicting. Iamblichus fr. 60.

Contents

Preface List of abbreviations

page ix xi 1

Introduction part i returning romance 1 First romances: Chariton and Xenophon

25

2 Transforming romance: Achilles Tatius and Longus

69

3 Hellenism at the edge: Heliodorus

108

part ii narrative and identity 4 Pothos

139

5 Telos

177

6 Limen

214 253

Conclusion Appendix: The extant romances and the larger fragments References Index

vii

261 265 295

Preface

This book marks the end of a long and serpentine journey. Versions of these chapters have been tested on audiences patient and inspirational in Atlanta, Berkeley, Birmingham, Bristol, Cambridge, Dublin, Exeter, Fresno, Geneva, Groningen, Leuven, Ljubljana, Lisbon, Liverpool, Los Angeles, Manchester, Michigan, Milan, Nottingham, Oxford, Paris, Pennsylvania, Princeton, Stanford, St Andrews, Swansea, Uppsala and Utrecht; I have also benefited immeasurably from the Welsh KYKNOS group (particularly meetings at Gregynog), the Cretan RICAN team led by Michael Paschalis, and participants in my own Romance between Greece and the East workshops. I have aimed to transliterate Greek names in their most familiar forms for ease of reading, accepting that no system of transliteration is perfect. Translations are mine, but I acknowledge my debt to other translators, especially those of Reardon (1989). For Achilles Tatius I have modified my own translations from Whitmarsh (2001b). For details of texts used please consult the appendix (divergences are noted throughout, where they occur). Iotas are printed adscript throughout; I have preferred bce/ce to bc/ad. I gratefully acknowledge the support of the Arts and Humanities Research Council, which (in its former life as a ‘Research Board’) awarded me research leave to allow me to complete a first version of this book. The Research Council also funded the Romance Between Greece and the East workshops alluded to above. Brill Academic Publishers generously allowed me to rework parts of Whitmarsh (2003) in Chapter 2. Chapter 3 contains material that originated in Whitmarsh (1998): thanks to the Cambridge Philological Society. Heartfelt thanks too, for multiple reasons, to Cliff Ando, Lucia Athanassaki, Siam Bhayro, Ewen Bowie, Camilla Chorfi, Irene de Jong, Elizabeth Dollins, Konstantin Doulamis, Ja´s Elsner, Dana Fields, Chris Gill, Maud Gleason, Simon Goldhill, Stephen Harrison, John Henderson, Owen ix

x

Preface

Hodkinson, Christopher Jones, Daniel King, Jason K¨onig, Rebecca Langlands, Anna Lefteratou, John Ma, Francesca Martelli, Stephen Mitchell, Silvia Montiglio, Helen Morales, John Morgan, Hannah Mossman, Karen n´ı Mheallaigh, Steve Nimis, Daniel Ogden, Boo Onion, Jim Porter, Ian Repath, Michael Sharp, Estelle Strazdins, Susan Stephens, Edmund Thomas, Stuart Thomson, Gail Trimble, Benet Walsh, and Froma Zeitlin. My gratitude to Francesca Stavrakopoulou, to my parents (Judy and Guy), brother (Ben), sister (Kate), and children (India and Soli) goes beyond words. Here’s to happy endings.

Abbreviations

ACM ANRW AP APM CA

DK FGrH GCN HU IE IG KA KAI3 LGPN LIMC

H. Musurillo, Acts of the Christian martyrs. Oxford, 1972. Aufstieg und Niedergang der r¨omischen Welt. Berlin, 1972–. Palatine Anthology. H. Musurillo, Acts of the Pagan martyrs. Oxford, 1954. J. Powell ed. Collectanea Alexandrina: reliquiae minores poetarum Graecorum aetatis Ptolemaicae 323–146 A.C. epicorum, elegiacorum, lyricorum, ethicorum. Oxford, 1925. H. Diels and W. Kranz eds, Die Fragmente der Vorsokratiker, 6th edn Berlin, 1951–2. F. Jacoby et al. eds Die Fragmente der griechischen Historiker. Berlin/Leiden, 1876–1959; continued Leiden, 1998–. Groningen Colloquia on the Novel. T. H¨agg and B. Utas, The virgin and her lover: fragments of an ancient Greek novel. Leiden, 2003. Die Inschriften von Ephesos, eds Wankel, H. et al. = Inschriften griechischer St¨adte aus Kleinasien 11. Bonn, 1979–. Inscriptiones Graecae, 2nd edn. Berlin, 1924–. R. Kassel and C. Austin eds, Poetae comici Graeci. Berlin, 1983–. H. Donner and W. R¨olling eds, Kanaan¨aische und Aram¨aische Inschriften. Wiesbaden, 1962–4. P.M. Fraser et al., A lexicon of Greek personal names. Oxford, 1987–. H. C. Ackermann and J.-R. Gisler eds, Lexicon iconographicum mythologiae classicae. Zurich, 1981–99. xi

xii LS LSJ MUSJ Pack

Pap. Mil. Vogl. P.Fay. P.Michael.

P.Oxy. P.Tebt. PMG PSI RG SVF SW TGF Wehrli

List of abbreviations A.A. Long and D.N. Sedley, The Hellenistic philosophers, 2 vols. Cambridge, 1987. H.G. Liddell and R. Scott et al., A Greek–English lexicon, 9th edn. with supplement. Oxford, 1996. M´elanges de l’Universit´e Saint Joseph. Beirut, 1906–. R.A. Pack, The Greek and Latin literary texts from Greco-Roman Egypt, 2nd edn. Ann Arbor, 1965; updated on the Mertens-Pack3 website http://promethee.philo.ulg.ac.be/cedopal. A. Vogliano et al. eds, Papiri della R. Universit`a di Milano. Milan, 1937–. Grenfell, B.P. et al. eds, Fayum towns and their papyri. London, 1900. D.S. Crawford ed. Papyri Michaelidae, being a catalogue of Greek and Latin papyri, tablets and ostraca in the Library of Mr G.A. Michailidis of Cairo. Aberdeen, 1955. The Oxyrhynchus papyri. London, 1898–. B.P. Grenfell, A.S. Hunt and J.G. Smyly eds, The Tebtunis papyri. London, 1869–1926. D. Page ed., Poetae melici Graeci. Oxford, 1962. Papiri greci e latini: pubblicazioni della Societ`a Italiana per la ricerca dei papiri greci e latini in Egitto. Florence, 1912–79. L. Spengel ed. Rhetores Graeci, 3 vols. Leipzig, 1856. H. von Arnim ed. Stoicorum veterum fragmenta. Leipzig, 1923–4. S.A. Stephens and J.J. Winkler, Ancient Greek novels: the fragments. Princeton, 1994. B. Snell et al. eds Tragicorum graecorum fragmenta. G¨ottingen, 1971–. F. Wehrli ed. Die Schule des Aristoteles: Texte und Kommentar, 2nd ed, 10 vols. Basel, 1967–9.

Introduction

true romance Habrocomes, my child, I am not a settler or a native Sicilian but an elite Spartan, from one of the powerful families there, and very prosperous. When I was a young man and enrolled among the ephebes, I fell in love with a citizen girl by the name of Thelxinoe, and Thelxinoe returned my love. We met at a time when an all-night festival was being held in the city, with a god’s guidance, and enjoyed the pleasure that our meeting had promised. For a while we used to meet in secret, and swore repeatedly to each other that our relationship would last till death. But one of the gods must have been spiteful. While I was still classed among the ephebes, Thelxinoe’s parents agreed to her marriage to a young man called Androcles, who had by this time also fallen in love with her. At first, the girl had to invent all sorts of excuses to put off the wedding, but in the end she was able to arrange a meeting with me, and she agreed to elope from Sparta with me by night. We both dressed up as young men, and I even cut Thelxinoe’s hair, on the very night before her wedding. Escaping from the city we came to Argos, then Corinth, where we boarded a ship and sailed for Sicily. When the Spartans learned of our escape, they condemned us to death. We lived out our days here, short of material comforts, but happy in the belief that we enjoyed every kind of pleasure, because we were with each other. Thelxinoe died here not long ago, but her body remains unburied: I keep it with me, maintaining my loving relations. (Xenophon of Ephesus 5.1.4–9)

The extraordinary story of Aegialeus the fisherman is one of a number of mini-novellas narrated by minor characters within Xenophon of Ephesus’ Anthia and Habrocomes, a Greek romance of the first century ce.1 It is 1

This introduction presumes a certain familiarity with the romances: for orientation, see the Appendix, where issues of dating are also discussed briefly. I use the term ‘romance’ for the heterosexual erotic narratives of travel and return, on which this book focuses, and ‘novel’ as a more extended category covering works like the Alexander Romance and The life of Aesop (both are, in any case, anachronistic).

1

2

Introduction

clearly an experiment with the romance mode. The themes of young, reciprocated, heterosexual love and adventure, fidelity and final happiness resonate with the primary narrative. This story of passion that survives beyond the florescence of youth, despite deprivation, is offered as a lesson in both the power of love and the harsh physiological and material realities of life. Habrocomes, Xenophon’s male protagonist and the recipient of this story, responds by drawing a conclusion that he applies to himself too: ‘now truly (al¯eth¯os) I have learned that true (al¯ethinos) love is not limited by age’.2 A true lesson about true love. But this story is also heavily counter-realistic. It is a grotesque parable about the delusions wrought by love. Aegialeus has (we learn) embalmed his wife in the Egyptian fashion, so that he can maintain the illusion that she is still living: ‘I speak with her constantly as if she were alive, and lie with her, and take my meals with her.’3 The final sentence of the story cited above (which I have translated ‘maintaining my loving relations’) could be taken to mean that he kisses and even has sex with the corpse.4 Aegialeus’ account of a life of poverty, exile, old age and death is not simply (as Habrocomes takes it) a story of true love; it is also about the denial of truth, about the concealment of present realities beneath a carapace of past memories. The lovers were, Aegialeus tells us, ‘happy in the belief (dokountes) that we enjoyed every kind of pleasure, because we were with each other’:5 this belief (doxa) that they remain prosperous is a fiction willingly entertained. Similarly, when Aegialeus proceeds to show Habrocomes her corpse, lovingly embalmed in the Egyptian manner, he tells him that ‘she does not appear to me as you see her; instead, my child, I imagine her as she was in Sparta, as she was when we escaped. I imagine the all-night festival, the promises we made.’6 Aegialeus seems neurotically obsessed with replaying his own teen romance, and adopting it as a substitute for reality. But it is not simply a case of false consciousness: he is fully aware that Habrocomes will see things differently, whereas he

2 3 4 5 6

For titles of the romances I use the girl–boy forms, which I believe to be original and generically definitive (Whitmarsh (2005b)); for convenience I abbreviate in the cases of Xenophon (full title: The Ephesian affairs of Anthia and Habrocomes) and Heliodorus (The Ethiopian affairs of Charicleia and Theagenes). Morgan (2004c) 491–2 discusses the relationship between Xenophon’s embedded narratives and his primary narrative; to his list I would add the story of Eudoxus at 3.4, reported in indirect speech. nÓn ˆlh{äv mem†{hka Âti ›rwv ˆlh{in¼v Âron ¡lik©av oÉk ›cei, Xen. Eph. 5.1.12. taÅthi . . . ˆe© te Þv zÛshi lalä kaª sugkat†keimai kaª suneuwcoÓmai, Xen. Eph. 5.1.11. filä (‘love’, but also ‘kiss’) kaª sÅneimi (‘conjoin with’), Xen. Eph. 5.1.9. ¡d»menoi . . . p†ntwn ˆpolaÅein dokoÓntev, Âti §men met ì ˆllžlwn, Xen. Eph. 5.1.8. oÉ g‡r o¯a nÓn ¾rtai soª toiaÅth fa©netai mo©, ˆll‡ –nnoä, t”knon, o¯a m•n §n –n Lakeda©moni, o¯a d• –n ti fugiá t‡v pannuc©dav –nnoä, t‡v sun{žkav –nnoä, Xen. Eph. 5.1.11.

Introduction

3

himself lives in a world of ‘belief’ and ‘as-ifs’. As a first-person (strictly, a homodiegetic) narrator, he stands both inside the story, living its fictions, and outside it, exposing them. Romance is centrally about simple truths: the complementary, yin–yang love of a girl and boy of the same station, comparable beauty and (roughly) equal age; a love tested through ordeals of separation and endurance, and redeemed through reunion and return. But literary narrative seems incapable of sheer simplicity. As we can see in the case of the story of Aegialeus (deliberately chosen from the romance usually reckoned the least artful), story-telling can be complex, self-conscious and metafictive even when it handles what is, at one level, a parable with an obviously universal relevance. This book is about identity in the Greek romances, and the ways that it is turned and re-turned through narrative. Identity is, of course, a hugely complex topic, spreading into history, philosophy, anthropology, psychology, sociology and cultural studies (particularly of postcolonialism, gender, race and sexuality).7 What I mean, for the purpose of this book, is primarily the set of categories of selfhood presumed, legitimised or questioned in the romances themselves. We can see immediately that Aegialeus uses a number of markers to identify himself. He is ‘not local’ to Sicily or a ‘settler’ (Sicelot, or Greek colonist), but an outsider, specifically a Spartan. He is a member of the elite (the Greek describes him as a ‘Spartiate’, of the city’s politically dominant class) and wealthy. Although the sentence expressing these claims omits the verb by ellipsis, the implication is that Aegialeus perceives this as a present-tense identity, which he still holds even in exile. He also refers, however, to transitory stages through which he has now conclusively passed: ‘when I was a young man . . . enrolled among the ephebes . . . classed among the ephebes’ (‘ephebes’ being males on the cusp of (epi-) maturity (h¯eb¯e)). A third mode of identity is the assumed disguise: ‘We both dressed up as young men, and I even cut Thelxinoe’s hair.’ These identities are provisional, strategic and designedly false; they will be shed when their usefulness is outlived. Finally, we have a less specific set of self-descriptors referring to mental and emotional states, principally the happiness generated by the illusory love. Even a brief story like this presents a rich narrative of identity. Aegialeus defines himself in terms of his city of birth, Sparta, but never achieved the secure status of adulthood there: he left while still an ephebe, not yet a man, just as Thelxinoe left 7

Discussion and references at Whitmarsh (2001a) 35–7; see now also du Gay et al. (2000), a sample of classic essays from a variety of fields.

4

Introduction

before she became a woman (i.e. wife). In place of their real, Spartan identities, they adopt first the false disguises they need for their escape, and second the consoling fictions that they are still the people that they were when they first met. There is a notable self-reflexivity to this narrative of identity: Aegialeus does not simply tell Habrocomes about his past, but also reveals the role that such story-telling plays in sustaining his fabricated world in the present. Narrative creates identities to inhabit in the present, as well as accounting for the past. That identity is a species of narrative is a truism in certain circles. Inspired by Paul Ricoeur’s monumental Time and narrative, Alasdair Macintyre’s After virtue, Charles Taylor’s Sources of the self, 8 and psychoanalytical critiques of the enlightenment identification of the person with consciousness, certain scholars have claimed that (to quote one) ‘the self, or subject [is] a result of discursive praxis rather than a substantial entity having ontological priority over praxis or a self with epistemological priority, as originator of meaning’.9 Even the social sciences, traditionally hostile to qualitative analysis, have caught the narrative bug.10 It is not my aim in this book to validate such ideas. My approach is historicist: I aim to show not what identity is (in a universal sense), but how it is configured within a particular body of literature. It happens that that body was (as we shall see presently) both durable and culturally central in the period under discussion, but narrative was certainly far from the only medium available to ancients for articulating and exploring identity. Numerous other media presented themselves (to name but a few: inscriptions, monuments, clothing, statues, coinage), which may have a narrative dimension, but are not constituted as narratives in any strong sense. Ancient theories of identity were numerous (principally from philosophers11 and medical writers, but we should include jurists too), but narrative does not play a central role in them.12 8 9 10

11 12

Ricoeur (1984–1988); Macintyre (1984), esp. 204–25; Taylor (1989), cf. esp. 47–8 (‘we grasp our lives as a narrative’, 47). On problems around the definition of ‘narrative’, see Ryan (2007). Kerby (1991) 4. For the general point, see Somers (1994), who argues that narrative studies offer better prospects for comprehending the perspectives of the dispossessed (see 613–17 on the social sciences’ rejection of narrative); also Polkinghorne (1995) and (more leisurely and epideictic) Bruner (1987). For an excellent study along these lines of narratives of motherhood, see Miller (2005). For a critique of the ‘psychological narrativity thesis’, see Strawson (2004), although his argument founders on the odd claim that episodic experience (which he opposes to diachronic) is not a form of narrativity. In chapters 5 and 6, we shall distinguish between ‘paradigmatic’ and ‘syntagmatic’ (roughly the equivalent of Strawson’s episodic) narrative. Gill (2006) focuses on Stoic and Epicurean ideas of selfhood, with plenty of lateral glances towards Galen (as well as Seneca, Vergil and Plutarch). Gill (2006) 69–73, on the minimal role of memory in ancient definitions of selfhood.

Introduction

5

Greek romances are not identity narratives in the sense that modern philosophers understand the term, which is to say articulations of individual selfhood. Certainly, we do find figures (like Xenophon’s Aegialeus) telling their own stories, sometimes at great length: one of the romances, Achilles Tatius’ Leucippe and Clitophon, indeed, is almost entirely narrated by the male protagonist. But, of course, such accounts are always embedded in larger narrative frames, which are themselves fictionalised. Leucippe and Clitophon is not an ingenuous attempt to express Clitophon’s identity, it is an experiment with a literary mode, building on a tradition of first-person narratives stretching back to Homer’s Odyssey. Yet, as we have begun to see in the case of Xenophon’s Aegialeus story, ancient romances do indeed encode paradigmatic models of identity, and have their own ways of theorising it. To understand what identity is doing in such texts, we need first to explore how narrative works in them, about the formative roles of genre and cultural context. inventing romance The Greek romance appears to have emerged in the first century ce, in Asia Minor. In antiquity it survived until at least the fourth century ce (whereafter it continued to influence poets such as Nonnus and Musaeus, as well as martyrologists and historians);13 it was later revived in mediaeval Persia and Byzantium.14 There are five texts that survive complete: from the first century, Chariton’s Callirhoe and Xenophon’s Anthia and Habrocomes; from the second century, Achilles Tatius’ Leucippe and Clitophon and, perhaps also, Longus’ Daphnis and Chloe; and, from the fourth century, Heliodorus’ Charicleia and Theagenes. Although in many ways different, each deals with a shared stock of narrative themes: the love of a young heterosexual couple, the trials that come between them, and a joyous reunion at the end. All are set in an imaginary, more or less classicising (i.e. Roman-less) world; Chariton’s and Heliodorus’ works are explicitly located in the classical period. In addition to the extant texts, we have a number of summaries by Photius, the swashbuckling ninth-century bishop of Constantinople, and an ever-increasing corpus of papyrus fragments that seem to share these concerns with young love.15 Some (like the fragmentary 13 14 15

Below, n. 61. For an up-to-date introduction to the Byzantine novels, see Burton (2008), with further literature. On the Persian version of Metiochus and Parthenope, see see n. 16 below. The most substantial collection of fragments and summaries is SW; all fragmentary romances are cited from there, unless otherwise stated. Five more fragments have been published in the interim: P.Oxy. 4760–2, 4811, 4945.

6

Introduction

romances of Sesonchosis, Ninus and Metiochus and Parthenope)16 are based on historical or pseudo-historical figures; others, such as Iamblichus’ Babylonian affairs and the Panionis fragment (P.Oxy 4811), are, like the extant texts,17 fictional. These romance texts should be seen against the back drop of a larger canvas of diverse novelistic literature of the imperial and even Hellenistic periods,18 including the Alexander romance, the Wonders beyond Thule of Antonius Diogenes, the Life of Aesop, Apollonius King of Tyre, the various Ass narratives,19 Lucian’s True stories, and the anonymous Joseph and Aseneth. Two themes distinguish the romance from other novels. The first is the reciprocated heterosexual love that we have already seen exemplified in Aegialeus’ story. The second is that of travel and return. This, I have argued, is conspicuously absent for Thelxinoe and Aegialeus: they are compelled to create ersatz identities because they do not return to assume their proper adult roles in Sparta.20 Why did the romance, this particular species of the ancient novel, emerge when it did, and why did it achieve such success? This question has occupied scholars since Pierre-Daniel Huet’s Lettre-trait´e de l’origine des romans, which argued that the romance was a west-Asian form that spread into Greek during the Hellenistic period.21 Erwin Rohde’s pivotal Der griechische Roman und seine Vorl¨aufer (1876), the founding work of modern scholarship in the field, is largely dedicated to retracing the Hellenistic Greek sources of the romance.22 But whereas older scholarship focused on producing narratives of diachronic development, critics since Perry’s The ancient romances (1967) have tended to emphasise the congruity between the romances and 16 17 18 19

20 21

22

For Metiochus and Parthenope, see HU, which includes as well as the Greek fragments and testimony an edition and translation of an eleventh-century Persian translation, ‘Uns.ur´’s V¯amiq u ‘Adhr¯a. Notwithstanding that in Chariton’s Callirhoe, the historical Hermocrates (Syracusan general at the time of the Athenian invasion) is father of the protagonist. Modern discussions of Hellenistic prose fiction: Ruiz Montero (2003); Whitmarsh (2010d). Scholarship has focused primarily on the triangular relationship between the pseudo-Lucianic Ass, Apuleius’ Metamorphoses and the lost Metamorphoses of Lucius of Patrae (Phot. Bibl. cod. 129): see esp. Perry (1967) 211–18; Van Thiel (1971); Mason (1994). The picture has been changed, however, by the publication of P.Oxy. 4762, a different version of the narrative (featuring, intriguingly, a ‘third-person’ (i.e. what narratologists call ‘heterodiegetic’) narrator). The implications of this have yet to be fully absorbed by scholars in the field. Comparably, Montiglio (2007) reads Apuleius’ Metamorphoses as a rejected return narrative. Huet (1670) 11: ‘l’invention [des Romans] et deu¨e aux Orientaux: je veux dire aux Egyptiens, aux Arabes, aux Perses, & aux Syriens’. More recent versions of the west-Asian argument, differently nuanced, can be found in Barns (1956), Anderson (1984), and Rutherford (2000); discussion at Stephens (2008). I shall address this issue in a forthcoming book, The romance between Greece and the East. ‘West-Asian’ is intended as a more neutral designation than the Eurocentric ‘nearEastern’. Other studies searching for literary origins: Lavagnini (1922) (local history); Giangrande (1962) (prose paraphrases); S. West (2003) (women’s tales).

Introduction

7

the cultural context of imperial Greece.23 These contextualising readings fall, broadly, into three different camps: (i) The first, emphasising the role of ‘private’ emotions and selfhood, sees the romance as the expression of a general reorientation away from the public sphere towards the inner person. Sometimes this is expressed in terms of a supposed transformation in civic culture: Greeks, it is claimed, had lost their sense of collective identity amid the vast territories of the Hellenistic and Roman worlds.24 A slightly different version of this interpretation, routed through the work of Paul Veyne and Michel Foucault, sees the romance in the context of an increased emphasis upon the ‘care of the self’: the elites under the empire, it is claimed, turned to self-discipline as a compensation for the political authority they had now lost.25 Yet another variety reads the romances as religious parables of virtuous suffering and redemption, expressing the world of the mystery cult.26 The usual assumption is that the texts were composed by and for elite males, but sometimes the emphasis upon emotional vulnerability is explained in terms of a demographically expanded readership, now incorporating the ‘bourgeoisie’.27 (ii) A number of critics have seen the romance as the product of a supposed reorganisation of sexual protocols in the imperial period. Foucault is influential here too, concluding the third volume of The care of the self with a brief chapter claiming the romances as articulations of a ‘new erotics’ of heterosexual mutuality, contrasting with the hierarchical phallocentrism of the classical period.28 They have been held to articulate the supposed centrality of marriage to the Greek aristocracies in the imperial period,29 the new prominence of women,30 and the identification of the sexual being with the innermost core of selfhood (a theme of Foucault’s own work). 23 24 25 26

27 28 29 30

In fact, Perry’s contextual analysis was already preempted by Rohde, who devotes a large section of his book to the ‘Second Sophistic’. Perry (1967), esp. 57–60; Reardon (1969) 293–4, (1991) 28–30; Morgan (1995) 143–7. Konstan (1994); MacAlister (1996); Toohey (2004). Ker´enyi (1927); Merkelbach (1962), (1988) (cf., implicitly, Petri (1963)). This view, which has not found general favour, is critiqued and/or nuanced by Turcan (1963); cf. also 1992), Geyer (1977), Stark (1989). See however Beck (2003) and Zeitlin (2008), who explore religious overtones more subtly. H¨agg (1983); Holzberg (1995). Similar claims have been made for the ‘Jewish novels’ of the Hebrew Bible: see e.g. Wills (1995) 3–6. Foucault (1990), followed by Konstan (1994); refinements in Goldhill (1995). This general approach is discussed by Morales (2008). See further below, pp. 159–60. Cooper (1996); Swain (1996) 101–31. Johne (1987), (2003); Egger (1988), (1994a), (1994b); Liviabella Furiani (1989); Wiersma (1990); Montague (1992); more circumspectly, Haynes (2003) 1–17.

8

Introduction

(iii) The third hypothetical context for the romances is the so-called ‘second sophistic’. When this term was (apparently) coined by Philostratus in the third century ce, it referred to a form of epideictic oratory, in which the speaker took on the persona of a figure from myth or history (VS 481, 507).31 When modern critics write of the ‘second sophistic’, however, they are usually making grander claims, about a supposed trend towards self-conscious Hellenic revivalism underpinned by the reinvention of links with the prestigious classical past. It was, again, the enormously influential Erwin Rohde who revived the phrase, linked it particularly with a supposed concern with the defence of ‘national-Hellenic’ (national-hellenisch) values against supposed eastern infiltration and Roman oppression,32 and located the romance within this supposed movement. Modern criticism tends to downplay the troubling distaste for ‘the East’ that Rohde’s model seems to both identify and endorse, putting the emphasis instead on anti-Romanism (so that the ‘second sophistic’ becomes a postcolonial rather than an anti-Semitic allegory).33 Against this, others have reinvented the second sophistic as a more playful, ‘postmodernist’ culture, revelling in its secondariness, self-awareness and sophistication. Here too, the romance has been seen as a prime exhibit, for its clever refashioning of traditional themes.34 We need to be careful here, since each of these contexts has its problems. As far as (i) goes, we can certainly point to the ability of romance narrative to go behind the scenes and portray emotions, but this kind of zooming technique is, in fact, as old as Homer. More problematic still is the belief that post-classical culture was mired in alienated ennui: this is little more than a modernist fantasy, and in some cases a teleological attempt to create a crisis for Christianity to resolve. There is no evidence for wide-scale anxiety, or for the collapse of polis culture.35 Inscriptions, monuments and literary sources (from Dio Chrysostom to Libanius) testify to the ongoing importance of civic culture, even if political structures were in flux. Conversely, public identity is extremely important to some of the romances, notably those of Chariton and Xenophon. The romances certainly contain expressions of despondency (as we shall see in chapters 5 and 6), but these are directed 31 32 33 34 35

This paragraph in part summarises the critique at Whitmarsh (2005a) 6–9; cf. also (2001a) 42–5. Rohde (1914) 319. See Bowie (1970) 9–10 on Chariton and Heliodorus; also Anderson (1993) 156–70; Whitmarsh (2001a) 78–87. See esp. Goldhill (1995), with ix on the second sophistic (and xi on their ‘wit, verve and outrageousness’); also, less directly, Morgan (1995) 142–3. Swain (1996) 106 effectively critiques the ‘anxiety school’.

Introduction

9

against fortune and malign gods rather than faceless world empires; and, what is more, they are almost always proven to be misguided. The romances do not demonstrate a shift from public to private identities, for no such shift had occurred. Instead, we should be focusing upon the question of how individual romances structure the relationship of private to public. Explanation (ii) – proposing a rise in conjugal ideology – is more helpful, since we can certainly point to an increased emphasis upon the representation of the virtues of marriage in a variety of media (epigraphy, literature and philosophy) from the late first century bce onwards,36 and a growing celebration of self-control, endurance and fidelity shared between Jewish, Christian and Greco-Roman cultures.37 On the other hand, it is once again far too simplistic to speak of a shift from hierarchical to symmetrical models of sexuality. Classical sexual protocols were not exclusively dominated by phallocentrism and power.38 Narratives of reciprocal heterosexuality redeemed in (the re-establishment of ) marriage are in the Greek tradition as old as the Odyssey, and lie at the heart of Hellenistic new comedy; marital devotion in the face of oppression is the theme of Xenophon’s celebrated narrative of Panthea and Abradates;39 ideas of sexual symmetry can be found articulated in classical mime.40 Conversely, there is plenty of evidence for asymmetrical pederastic desire in the imperial period.41 The picture that emerges is of subtle adjustments in a complex system, rather than of a sudden, decisive break. Explanation (iii), the ‘second sophistic’, also has its difficulties. There is no doubt that the romances (with the exception of Xenophon’s Anthia and Habrocomes) are highly sophisticated products of elite, educated Greek culture.42 They are also composed in a prose (a hallmark of imperial Greek 36

37 38 39

40 41 42

See van Bremen (1996) on the epigraphic record. Milnor (2005) links the reorientation closely to Augustus; see esp. 239–84 on the centrality of marriage to philosophers such as Musonius Rufus (see also Whitmarsh (2001a) 109–13; Nussbaum (2002)). See also Swain (2007) 146–52 on the intriguing Bryson, who survives only in Arabic translation. The classic statement of this position is Brown (1990a), summarised at (1990b). Perkins (1995) integrates the romances into her study of the ethics of endurance. Davidson (2007) is controversial, but on this point (I think) absolutely right. Xen. Cyr. 4.6.11, 5.1.2–18, 6.1.45–51, 6.4.1–11, 7.1.29–32, 7.3.2–16. The importance of this narrative for the romances is well-known: see most recently Capra (2009) on Xenophon. The Panthea story was rewritten (perhaps as a rhetorical novella) in the second century ce by one Celer (Philostr. VS 524). Xen. Symp. 9.6: ‘the boy and the girl are kissed by each other’ (t¼n pa±da kaª tŸn pa±da Ëp ì ˆllžlwn file±s{ai). Below, p. 160. On the evidence for elite readership, see S.A. Stephens (1994); also Bowie (1994). Cavallo (1996), by contrast, argues on papyrological grounds that the reading public for the romances diversified in the second century; but his argument depends heavily upon judgements as to what ‘un lettore non abituato a testi di cultura superiore’ (35) would expect from a text.

10

Introduction

aesthetics)43 with marked Atticising tendencies (in the cases of Achilles, Longus and Heliodorus), and their intertextual reference points are broadly in line with those of other imperial authors.44 But whatever it was (and I am increasingly sceptical that it was anything very much), the ‘second sophistic’ was not a unified, manifesto-led organisation. If the romances share some features with other literary productions of the era, it does not follow that they are entirely of a piece with them. We should be particularly careful about Rohdean claims that the era was dominated by a stridently defensive Hellenism. The disturbing political implications apart, it posits a wholly implausible uniformity across a huge time and space, in an age long before nationalist mechanisms like print media. I shall argue in the course of this book that Chariton and Xenophon do display a Hellenocentrism of a kind, subtle and complex in Chariton’s case. Achilles Tatius and Longus, however, configure identity very differently. Iamblichus and Heliodorus, finally, offer direct challenges to the Hellenocentric model.45 What is crucial is to get away from the paradigm shift model. Relationships between historical processes and the invention of cultural forms are, as a rule, complex and multiform. In some cases we can certainly point to social or political events that impel new genres: for example, the ‘May 4th Movement’ of 1917, which created the conditions for the rise of the vernacular Chinese novel.46 More typically, however, literary works are shaped by multiple influences, which may include, alongside social, political and cultural shifts, the conservatising effects of canons and traditions as well as the idiosyncratic creative aspirations of individual authors. Scholars of Greek tragedy, for example, have retreated from the dogma that the genre was entirely shaped by Athenian democracy. There are, of course, democratic resonances in the interplay between named individuals and anonymous collectives, the relativisation of authority, and the emphasis upon the fall of royal households. But sceptics are right to point out that such themes are already found in literature predating Athenian democracy,47 and indeed that they are ‘civic’ rather than narrowly democratic.48 Democracy may be a necessary cause of the emergence of tragedy, but it is not a sufficient one: a full account would need also to address other genealogies of the genre, for example, in Dionysiac ritual, epic narrative and choral festivals. 43 44 45 46 48

On the prosiness of imperial Greece, see Whitmarsh (2005c). For general overviews see Fusillo (1989) 17–109; Morgan and Harrison (2008) 218–27. For Heliodorus’ ‘multiculturalism’ see Bowersock (1994) 29–53; Whitmarsh (1998), (1999); Perkins (1999); below, chapter 3. Zhao (2006), esp. 83–6. 47 Griffin (1998), esp. 48–9. Rhodes (2003), titled ‘Nothing to do with democracy’; also Taplin (1999) on tragedy’s trans-civic portability.

Introduction

11

The emergence of romance is even more difficult to relate to particular historical changes, because it was, it appears, composed outside of civic institutions. At least with tragedy we can be sure that the Great Dionysia was sponsored by the Athenian democratic state, even if not all tragedies were composed for that festival exclusively (or, in some cases, at all), and even if the performative context did not fully dictate the form that the texts took. The romancers, by contrast, wrote for readers and environments that they could not predict and would never encounter. Their works were composed in the awareness that they would travel beyond the immediate community in which they were composed, to the literate elite across the empire: hence, for example, the appearance of the Aphrodisian Chariton in rural Egypt some 100–150 years after the text’s composition (and, for all we know, even earlier as well).49 The written word also engenders a greater sense of the longue dur´ee of literature. A millennium or so of literate culture creates not only a rich sense of the past (the romances are all richly intertextual50 and set, whether specifically or not, in a yesteryear), but also the expectation that one writes for posterity too.51 When Longus claims that he has written a ‘pleasurable possession for all humanity’,52 it is more than just an allusion to Thucydides. Now, of course, this universalism can itself be situated historically, albeit in a rather general sense: it is clearly the product of an expanded Greek worldview (from the Panhellenism of fifth- and fourth-century Athens, through the Hellenistic period, to Roman imperialism), the material opportunities of the pax Romana, and the culture of classicism. It is attractive to see the ancient romance as fundamentally shaped by new modes of literate production and transcultural circulation.53 Walter Benjamin famously claims that the modern novel is born of print technology; its appearance marked the end of face-to-face communities and the birth of the ‘solitary individual’ of deracinated consumerism.54 It would be misleading to suggest too close an analogy (even if we were to accept Benjamin’s judgemental distaste for the reproduction and circulation of texts), but it is not 49

50 51 52

53

P.Oxy. 1019 and P.Oxy. 2948 (from the same papyrus = 241 Pack), P.Michael. 1 (242 Pack), P.Fay. 1 (243 Pack), all dated to approx. 200 ce. The codex Thebanus deperditus (244 Pack) is dated to the sixth to seventh centuries ce. Above, n. 44. Estelle Strazdins’ forthcoming Oxford DPhil will demonstrate the importance of this motif in imperial Greek culture. ktma . . . terpn¼n psin ˆn{rÛpoiv, 1 pr. 3. The universalism is formally contrasted, via a men/de construction, with the metaphor of an epichoric cult dedication (ˆn†{hma . . . ï Erwti kaª NÅmfaiv kaª Pan©). Along these lines, see e.g. Morgan (1995) 137–8. 54 Benjamin (1970) 87.

12

Introduction

far wrong to see the ancient romance as similarly emerging from the literate interconnectivity of the Hellenistic and (particularly) Roman empires. I return to this point in my conclusion. But if the appearance of (aspirant) ‘world empires’ created the conditions for the romance to arise and prosper, historical determinism still helps little with the content of heterosexual love, ordeals and travel – themes that are, by contrast, widespread in all periods of Greek culture (see the following section). Nor does it explain the specific thematic inflections of the different authors (for example, Longus’ pastoral context or Achilles Tatius’ parodic approach). This, of course, is not at all the same thing as saying that the individual romances or their components float free of historical anchorage. It is a central claim of this book that these texts respond, diversely, to sociocultural exigencies. The crucial point, however, is that we should not see the relationship of history to genre as one of cause and effect. Each of the romances is itself a distinctively creative working-through of contemporary identity politics. returning romance What, in fact, do we mean by referring to romance as a ‘genre’? The romance is, for many, the limit case of generic fluidity.55 ‘The novel becomes the great container in which reified fragments of these genres will take their place alongside others’;56 far from a genre in its own right, it is constitutively ‘anti-generic, unable to be specified as a single style of discourse; it is a container of styles rather than itself a homogeneous and distinctive style’.57 This impression of genrelessness is seemingly reinforced by the absence of any evidence for any institutionalisation of the romances: they were not, so far as we know, performed;58 nor did they appear in school curricula or the 55

56 57 58

Farrell (2003) 391–3. See Holzberg (2003) 11–16 for a survey of attempts to define the genre of the ancient romance; also Goldhill (2008). In scholarship of post-classical romance, the denial of genre has become routine: see e.g. Parker (1979); Elam (1992) 5–8. Nimis (1994) 407. Nimis (1994) 398; for similar points, see e.g. Fusillo (1989) 26; Harrison (2003) 515. For a critique of this kind of reading of Bakhtin, see Whitmarsh (2005d). There is, however, some evidence for theatrical performance of the stories we know as Metiochus and Parthenope (Luc. De salt. 2, 54; Pseud. 25) and Ninus (Luc. Pseud. 25), but in the form of mimes rather than recitations of the romance text: this is all but explicit at De salt. 2 (see further HU 49–52). The relationship between romance and pantomime may have been bilateral: see Mignogna (1996a) and (1997) for mimic themes in Achilles Tatius, and (1996b) for a proposed identification of a fragmentary papyrus as a Leucippe mime. Among several speculations, Obbink (2006) wonders whether P.Oxy. 4762 might be a mime version of the ass story. To re-emphasise, however, there is no evidence that the romances themselves were performed. More generally on romance and mime,

Introduction

13

reading lists of advanced literary criticism.59 There is not even any ancient word for ‘romance’: even the most plausible candidate, dramatikon, does not appear before the ninth century, and even then seems to refer to the ‘dramatic’ aspects of the plot (sufferings and reversals of fortune) rather than a genre itself.60 Nor, finally, is there any explicit theoretical discussion of the Greek romances before Byzantine times (although allusions in other texts61 indicate that they were certainly being read).62 The characterisation of the romance as an anti-genre is, I think, too strident;63 in fact, it only works if we define genre in an antiquated, essentialist way. Genres are in general, as constructionist scholarship has taught us, fluid things, liable to continuous reinvention; they are structured by

59

60

61

62

63

see Webb (2008) 27, 96–7, 136; and on mimes and pantomimes as a whole Rouech´e (1993) 15–30, Lada-Richards (2007), Hall and Wyles (2008) and Webb (2008). Not that this is a surprise, since imperial Greeks rarely discuss their contemporaries (Bowie (1994) 442). Genette (1992) 69: ‘the postclassical (or paraclassical) forms suffer a historical erosion that is less their own doing than that of another historical rhythm’. Philostratus is barely mentioned (Menander Rhetor 2.390 RG = Russell and Wilson (1981) 116; and in his homonymous grandson’s Imagines (pr. 2)), as is Lucian (Strohmaier (1976) for the reference in Galen, transmitted in Arabic; Lactant. Div. inst. 1.9; Eunap. VS 454). For dramatikon used of romances, see Phot. Bibl. cod. 73, 50a = Hld. test. IV Colonna; cod. 87, 66a = Ach. Tat. test. 2 Vilborg; cf. cod. 166, 109a (Antonius Diogenes). But in fact Photius happily uses the term of works that are not romances: see e.g. 95, 78b; 107, 87b. For discussion of this term see Rohde (1914) 376–9; Agapitos (1998) 128–30. Nor does plasma (‘fiction’), the word used in the colophon of the earliest MS of Achilles Tatius (W, twelfth century) have any secure pre-Byzantine ancestry, despite Jul. Ep. 89 = 301b (this does not refer to the romances: see Whitmarsh (2005b) 607–8). In any case, this word, which derives from rhetorical theory (see esp. Barwick (1928)), refers not to the literary form but to the type of scenario described, i.e. invented as opposed to historically or mythologically based. Similarly, er¯otika (‘erotic things’), er¯otikon di¯egema (‘erotic narrative’) and er¯otik¯e hypothesis (‘erotic plot’) are simply bland descriptions of the contents. The variety of terms is further discussed at Kuch (1989) 13–14. I argue at Whitmarsh (2005b) that the romance’s titling conventions (ta peri/kata + girl’s name) are used to denote genre. See esp. Kost (1971) 29–32 on Musaeus. Lehmann (1910) contains a useful, if not always convincing, list of ‘imitations’ of Achilles in Aristaenetus (5–12), Musaeus (12–25), Synesius (25–6), Gregory of Nazianzus (28–30), Himerius (30–6), Themistius (36–9), pseudo-Libanius (39–46), Nonnus (46–7) and Philostratus (51–4). Julian’s 89th letter does not, contrary to what is often claimed, refer to the romances: see n. 60. A number of other references have been claimed: I am unconvinced by Persius 1.134 (post prandia Callirhoen do: see Whitmarsh (2005b) 590 n. 14) and, while Philostr. Ep. 68 is a more plausible reference, it is too brief and allusive to allow any certainty. The first secure mention of a romance is the famous reference to Heliodorus’ supposed bishopric by the fifth-century Socrates Ecclesiasticus (HE 5.22 = T I Colonna). The Latin novels (explicitly including Apuleius), however, are discussed as a genre (hoc . . . fabularum genus) by Macrobius in the late fourth or early fifth century (Somn. 1.2.8, with Kuch (1989) 13–16). That the Greek romances, to return to them, were read in relatively (though not spectacularly) large numbers is evidenced by papyrus finds (up to forty-two in 1994, depending on what one counts: see S. A. Stephens (1994) 415–16), the influence on authors such as Musaeus and Nonnus, and the mosaics from Roman Syria representing scenes from the narratives (if perhaps the mimes rather than the romances: above, n. 58) we know as Ninus and Metiochus and Parthenope (HU 57–64). Whitmarsh (2005b); (2005d) 108–11.

14

Introduction

‘family resemblance’ rather than phylogenetic classification.64 In the case of the romances, shared themes, topoi, and language make it certain that these writers were self-consciously constructing their own relationship to a tradition.65 The crucial point is to think of genre not as an ossified ‘form’66 but as a flexible system allowing for reinvention and reorientation. Part I of this book aims to demonstrate this. Three chapters describe three phases in the history of the romance, belonging to the first, second and fourth centuries ce. Broadly speaking, the earliest romances celebrate the regenerative power of the Greek community; their second-century successors are more centrifugal and experimental, emphasising the flexibility and capacity for transformation that experience can supply; while Heliodorus’ fourth-century Charicleia and Theagenes constitutes a radical reorientation, relocating the idealised marriage-based community from Greece to the edges of the earth. What this book thus offers is an account of a literary form that was both traditionally rooted and flexible enough to respond to changing circumstances. The defining feature of the romance plot is, I argue, the return narrative. The return (nostos) is among the most ancient and fundamental in the Greek world (perhaps with its roots in ancient West Asia). Homer’s Odyssey is the best-known example of this type to Hellenists, not just for Odysseus’ literal return, but also for the counterfactual return narratives contained in the hero’s various ‘Cretan lies’, and the gossip told to Penelope about him;67 such tales of travel and return were no doubt legion in the ancient Mediterranean (much earlier examples include the Egyptian Sinuhe and the Tale of the shipwrecked sailor, from the second millennium bce). Their popularity in Greece can be glimpsed from the remains of the Nostoi of Agias of Troezen (which told of the returns of the Achaean warriors at 64

65

66 67

The Wittgensteinian phrase comes from Fowler (1982). See further Heath (2004) 168. ‘A genre . . . is not simply a taxonomic class . . . If a theory of genres is to be more than a taxonomy it must attempt to explain what features are constitutive of functional categories which have governed the reading and writing of literature’ (Culler (1975) 137). For a thoroughly conventionalist reading of genre, see esp. Genette (1992), a revision of Genette (1977); also Derrida (1980), arguing that genre exists only as a call to an impossible purity. On ‘speech genres’ see esp. Bakhtin (1986) 132–58, lucidly discussed by Morson and Emerson (1990) 271–305. Seitel (2003) has some stimulating remarks on genre as speech act from an anthropological perspective; see also Pavel (2003). See esp. L´etoublon (1993) on lieux communs. One of the results of researching this book has been the realisation of how much intertextuality there is within the corpus of romances. I have noted instances where they have arisen, but a systematic study of this phenomenon is a desideratum. Reardon (1991). On Odysseus’ nostos as identity narrative, see esp. Segal (1962) and (1967); also Goldhill (1991) 1–68, emphasising the roles of language and representation. Frame (1978) argues that Odysseus’ return embodies a a death/rebirth scheme. On the wider nostos tradition, see Alexopoulou (2006), (2009).

Introduction

15

Troy, and Orestes’ vengeance on Aegisthus and Clytaemnestra) and of Stesichorus (PMG 209),68 and various tales in the wider mythic corpus (such as those of the Argonauts and Heracles, as well as the later Attic mythology of Theseus). Return stories are typically optimistic and joyous, figuring the renewal of society, and even life itself: hence their alignment with stories of return from the underworld (Heracles, Odysseus,69 Theseus)70 and the joyous restoration of the divine order and hence of fertility (the ‘return of Hephaestus’ to Olympus, featured on a number of archaic black-figure pots,71 the Homeric Hymn to Demeter,72 the Mesopotamian Nergal and Ereshkigal, and Egyptian accounts of Isis and Osiris – to which Ker´enyi attributed the origin of the Greek romances).73 It has been argued that the etymology of the very word nostos suggests a ‘return to the light’, and that early Greek nostoi like the Odyssey manipulate rebirth symbolism.74 Crucially, return narratives also operate as parables of identity, normative articulations of relationships between self and society: they imply the destiny of the individual (human or divine) in a particular context. Underlying such stories is often a ritual pattern, which scholars (following Van Gennep) tend to call the ‘rite of passage’: common to many pre-industrial cultures is the idea that the individual is aggregated into society after a period of marginalisation.75 What romance distinctively superimposes onto the common return plot is the narrative of heterosexual desire. Greek romances, without exception, begin with the excitation and conclude with the satisfaction of such desires.76 This combination is not without precedent in the Greek 68

69 70 71 72

73 76

A Nostoi is also ascribed to Eumelus of Corinth (S Pind. Ol. 13.31a), an ascription that may, however, ‘be an isolated error’ (M. L. West (2003) 26). The archaic nostoi are surveyed by Malkin (1998) 1–10, with a particular emphasis upon their role in the colonial imaginary. Neoplatonists read the Odyssey as an allegory of the soul’s return from the phenomenal world to the pure world of forms: Lamberton (1986) 106–7. Frye (1976) 97–126 observes the proximity between romance and katabasis/anabasis. Hedreen (2004), with literature. The significance of the Hymn to Demeter as a canonical text of yearning, separation and ‘rebirth’ may suggest that it is more than coincidence that the list of Nereids at HHomDem 418–23 contains names that would later belong to three romance heroines (Leucippe, Callirhoe, Melite (already a nymph at Hom. Il. 18.42 and Hes. Th. 247); Rhode (cf. Rhodeia, HHomDem 419) is also the name of a slave in Xen. Eph.). The Europa episode at the beginning of Achilles Tatius seems to allude to this passage (among others). Ker´enyi (1927). 74 Frame (1978). 75 See below, pp. 43–4. One possible exception is Metiochus and Parthenope, which HU 247–50 claim may have had an unhappy ending. I am unconvinced by their arguments (that Parthenope’s name suggests perpetual virginity; that there is no trace of reunion in the fragmentary testimony; and that St Parthenope was martyred), but agree that the case remains open.

16

Introduction

tradition, although the parallels are inexact. In Homer’s Odyssey, the protagonist’s yearning for homecoming is merged with his desire for his wife, a desire that blends the sexual and the social.77 Aristophanes’ Lysistrata begins with the secession of women from their ‘proper’ place in the community, and concludes with the reincorporation of the genders, symbolised by the communal dancing of female and male choruses. Many a New Comedy focuses on sexual crises that are finally resolved by integrating the sexes into the married household.78 Comic sex plots, however, tend to avoid travel, and hence the nostos. Romance is different: partly because of the medium in which it is circulated (the prose book), but more pertinently for our purposes because it combines the narrative of desire fulfilled with the rite of passage/return plot. In other words, the protagonists begin the narrative as yearning youths and end as fulfilled adults. Let us recall, in this connection, that Aegialeus’ story begins ‘when I was a young man and enrolled among the ephebes.’ Heterosexual marriage is, in the romances, the primary marker of the achievement of adult identity in the polis (or not, in Aegialeus’ case). re-turning romance The romances, however, are not simply normative maps of identity. Certainly, they can be read in this way: one reason for the very persistence of the return narrative in general is that it responds to a deeply rooted human need to naturalise the human subject, with all his or her uncertainties and desires, within a community: to engender a powerful sense of home and homeland as the telos of existence. But home-coming raises questions too: is the returner the same person as the one who left? As Terence Cave argues in a sparkling reading, scenes where home-comers are recognised always (from the Odyssey onwards) seem haunted by the Martin Guerre scenario, in which an imposter is falsely acknowledged.79 The romances respond to this anxiety. As Cave himself observes, Heliodorus’ Charicleia and Theagenes contains a protracted, forensic analysis of Charicleia’s identity, so as to determine that she is the true child of the Ethiopian royal couple, with Hydaspes (her father) particularly sceptical.80 How can we be sure she is who she says she is, he asks (10.13.5)? Different versions of 77 79

80

Below, p. 142. 78 See e.g. Hunter (1985) 83–95; Lape (2003). Cave (1988). The French peasant Martin Guerre left home in 1548. In 1556 a man claiming to be Guerre returned and was welcomed by his wife; it later transpired that he was an imposter who had campaigned with Guerre. Cave (1988) 17–21.

Introduction

17

this question arise in Lockean form even in cases where the identity of the returner is not explicitly called into doubt. In Chariton’s Callirhoe, the heroine and Chaereas are welcomed back to Sicily by joyous crowds in the final book. But are they the same as the people who left? Callirhoe has learned that deception is the secret of a happy marriage; Chaereas has learned to redirect his vigour from wife-beating to militarism.81 Their experience has transformed them. At the same time, however, the return home always suggests the restoration of a prior state, as though the interlude had changed nothing. Bakhtin claims that there is no ‘biographical time’ in the Greek romance: ‘At the novel’s outset the heroes meet each other at a marriageable age, and at the same marriageable age, no less fresh and handsome, they consummate the marriage at the novel’s end.’82 This is (as we have seen) not true in an absolute sense, but it captures an aspect of romantic sameness. The protagonists are not ravaged by age and experience like Odysseus; all the sufferings inflicted upon them during the course of the narrative are always reversible. As Frye observes, the Greek romances’ emphasis on the integrity of female virginity works as an allegory for the immutability of the self despite everything.83 There is also a sense of the dependable eternality of the community, at least in the earlier texts: nothing changes in the polis either. Returning romance enacts a paradox: a text in which the protagonists both are and are not the same at the end. This paradox would, I think, be more familiar (but not necessarily more resoluble) for members of a community with a strong conceptualisation of passage rites. Such rites are often imagined as forms of death and rebirth:84 so when initiates are welcomed into the community, they are in one sense simply the child now fulfilled, and in another a different person. A late Greek orator describes the experience of mystery cult initiation as ‘becoming a stranger to myself ’,85 a baffling phrase marking the complex duality of this initiate, both within and without his old self. 81 83 84

85

Balot (1998); Scourfield (2003). 82 Bakhtin (1981) 90. Frye (1976) 86. See also below, pp. 144–5. Already at Plut. fr. 178 Sandbach, esp. lines 6–7: ‘the language and the reality of death [teleutan] resemble those of initiation [teleisthai]’ (t¼ çma täi çžmati kaª t¼ ›rgon täi ›rgwi toÓ teleutn kaª tele±s{ai pros”oike); and see e.g. Turner (1967) 96 on the Ndembu. Greek marriage in particular is often linked to death, especially for the parthenos (maiden) who will be reborn as a gun¯e (woman). See esp. Rehm (1994), Seaford (1987). In Heliodorus, Charicles’ first daughter does actually die on her wedding night (2.29.3–4), a widespread topos mapped out by Szepessy (1972); Knoles (1980–1) traces it beyond antiquity. xeniz»menov –pì –mautäi, Sopater 114.26–115.1 Walz; see below, p. 101.

18

Introduction

It is this unresolved play between sameness and difference that gives romance narrative its richness, complexity and urgency. Todorov once argued that such issues are the very essence of narrative itself. ‘Narrative’, he claims, ‘is constituted in the tension of two formal categories, difference and resemblance’, since the journey between beginning and end must describe both continuity and change: transformation represents precisely a synthesis of differences and resemblances, it links two facts without their being able to be identified. Rather than a ‘twosided unit’, it is an operation in two directions: it asserts both resemblance and difference; it engages and suspends time in a single movement . . . in a words, it makes narrative possible and yields us its very definition.86

Whatever its merits as a general theory of narrative, this nicely summarises the experience of reading romance. This doubleness has multiple ramifications, which will be traced throughout this book, particularly in part II. There I explore the problematic, challenging nature of the returning romance. Why is it that the same text can be read as conservative expressions of traditional family values, or as a sexual phantasmagoria; as a closed, teleological form or as an open-ended experimental; as Hellenocentric or as centrifugal; as philosophically serious or as comedic? This is in part a question of which elements we choose to emphasise, for these are compendious texts containing many different kinds of utterance, often contradictory. In Heliodorus, for example, we find the same character (Charicleia) variously referring to the cosmos as chaotically unstructured and providentially ordered.87 One central aim of part II is to interpret such perspectives relationally. As scholars of post-classical literature have emphasised, romance plays out conflicting desires, particularly in relation to closure. For Patricia Parker, notably, it is ‘a form that simultaneously quests for and postpones a particular end’.88 For David Quint, epic and romance are opposed as teleological linearity to digressive episodicity; but in any given text, the two forces will operate dyadically, the one always implying the other.89 Following such scholars (and this is part of the reason for my preference for the term ‘romance’ over the more conventional ‘novel’), I read the Greek romances as complex and conflicted, at the level of metanarrative, which is to say a text’s self-conscious theorising of its own narrativity. There is much 86 88

Todorov (1977) 233. 87 Chaotic: 6.8.3–6; providentially ordered: 9.24.4. Parker (1979) 4. 89 Quint (1993), esp. 31–41.

Introduction

19

about narrative in this book, but I am not conceiving of this in the static, definite sense familiar from the formal narratology that has dominated the field (much though I have learned from this).90 This book treats narrative instead as a space of indeterminacy, for the characters of romance and for the reader alike, at least until the closural resolution (and, to an extent, even then). In this respect, I build on the cognitive approaches of critics like D.A. Miller and Peter Brooks, who treat narrative as a realm of possibility and desire.91 Narrative, from this perspective, is not simply a synonym for Aristotelian plot – the system that gives events their larger significance through logical or plausible concatenation – but rather an experiment with multiple, sometimes contradictory plotting hypotheses. It may seem surprising to speak of conflicted desire in the romances, whether at the literal or the metanarrative level. Aegialeus and Thelxinoe are the exceptions: elsewhere, almost all other (sympathetic) characters want to go home and live a stable life of marital happiness with their partners. If these articulations of desire figure or stimulate the reader’s own sense of plot, can we not simply conclude that romance is entirely trammelled towards the happy ending? At one level, this is surely right: the point bears repeating, that return stories play to a deep, structural desire to locate the individual in a larger pattern of optimistic renewal of life through communal living. As the Aegialeus story has already demonstrated, however, the Greek romances are not simply reflexes of primal instincts, but also mediated through artful narrative. Even by the early imperial period, Greek writers could call upon some 800 years’ worth of sophisticated return narratives, together with critical reflections upon them. To give a full account of the romance, we need to explore also the turns of narrative and language, the tropes (tropai) that are so fundamental to Greek adventure narrative, from polytropic Odysseus onwards. The book is not just about the return romance as paradigm; it is also about how individual romancers cunningly re-turn the tradition. Even more than any narrative form,92 romance needs procrastination, unpredictability and amplification, and for a number of related reasons. The first is obvious, namely that in a return narrative, the plot exists 90 91

92

Egregious examples include H¨agg (1971), Lowe (2000) 222–58, and Morgan (2004b–e) and (2007b), (2007c), (2007d), (2007e); see also Whitmarsh and Bartsch (2008) 237–45. Miller (1981); Brooks (1984). See also Sturgess (1992) on narrativity, ‘the enabling force of narrative, a force that is present at every point in the narrative’ (28); Sturgess too conceives of narrativity as potentially contradictory, capable of sustaining ‘double logics’ (68–92). Lowe (2000) 65–8.

20

Introduction

primarily as a detour. The limit case of a romance with no turns would involve its protagonists staying at home throughout; in other words, it would lack all romance narrativity. The second reason follows from the first: the travels introduce the difference into the identity narrative. If I leave the house to buy a newspaper and return five minutes later, I am substantially the same person when I return. If, on the other hand, I am abducted by slavers and return only after three years, I will be transformed. In romance, the travels are the location for what Derrida would call diff´erance: a deviation, both temporal and spatial, from the linearity that constitutes identity (in its root sense of sameness).93 Finally, it is during the detour that romancers get to express their artistry: not just in the ecphrastic descriptions, imported from sophistic oratory, that intercut and arrest the narrative flow, but also in the very creative exuberance that contrives ever new twists and turns. When Chariton refers to Fortune as ‘keen on invention’ (philokainos, Char. 4.4.2), for example, he is self-reflexively vaunting his own capacity to fashion surprising new episodes.94 So even if the principal characters express desire for a speedy return home, there are other desires articulated, albeit more subtly, for centrifugal expansiveness. It is in this sense that romance narrative is conflicted. In the course of this book, I argue that the identity modelled in romance incorporates this conflict. In Chapters 5 and 6, I explain this identity in terms of the Freudian division between ego, superego and id. The ego is the ‘realist’ level of narrative, experienced as a mimesis of the real world: its consciousness, as it were. The superego is the desire for home-coming, and for all the conservative ideological apparatus thereby implied (marriage, social and familial role playing, hierarchies of gender and class). The id, by contrast, is the realm of emotional turbulence, centrifugality, narrative polytropism, alterity, the transformation of identity.95 The use of this terminology is an analogy (that is to say, it is not intended to imply the universality of Freudian psychology in a literal sense); but it is a productive one, because it allows us to extrapolate from the complexity of romance narrative a more sophisticated model of identity, one that transcends simple either/or models (conservative/experimental, Hellenocentric/Hellenofugal, public/private and so forth). It is, finally, this flexible model of narrativity and identity that, I argue, accounts for the success of the romance over a period of three hundred or 93 95

For diff´erance as detour, see esp. Derrida (1974) 85–7. 94 Below, p. 247. Fludernik (2007) 264: ‘In a literal reading of the romance quest motif . . . the other is a space of alterity . . . From a psychoanalytic point of view, these uncanny spaces of alterity symbolize the unconscious or id.’

Introduction

21

so years (in antiquity alone). It means that individual romancers can offer a satisfyingly rich and multifaceted narrative, capable of accommodating divergent readers with their own perspectives and tastes. But it also means that the form is malleable enough to respond, over time, to new historical and cultural circumstances, as different authors adjust the transformational ratio. It is to this diversity that we turn in part I.

part i

Returning romance

chapter 1

First romances Chariton and Xenophon

The Greek romance as we know it seems to have achieved its canonical form in the first century of the Roman principate.1 For sure, Hellenistic precedents may well have existed, particularly in the ‘national literature’ of the subject peoples of the Greek kingdoms:2 one particularly important case is Joseph and Aseneth, which tells of the mutual love, marriage and tribulations of the biblical patriarch, but the dating remains controversial (estimates vary between the second century bce and the fourth century ce).3 But it remains true, on the current consensus, that the ideal, fictional romance as we know it is very much a product of the early imperial era. The two first-century romances that survive in full4 are Chariton’s Callirhoe and Xenophon’s Anthia and Habrocomes. Each offers a broadly similar narrative: young lovers meet, fall in love, are separated abroad, and return to be reunited. Each is set in a major Greek homeland (Syracuse for Chariton, Ephesus for Xenophon); the temporal setting is, implicitly (Xenophon) or explicitly the past (Chariton locates his romance in the aftermath of the failed Athenian invasion of 415–413 bce; his heroine is the daughter of the historical general Hermocrates). How can we explain the emergence of this genre at this particular historical juncture? In my introduction, I argued that we should not look to historical determinism alone to explain the emergence of the romance; we should be thinking of how individual authors creatively construct paradigms of identity rather than expecting them to reflect them passively. Over the course of this chapter, we shall see how Chariton and Xenophon exploit 1

2 4

It was once thought that Chariton belonged to the first century bce, and the Ninus fragment was pushed back as far as the second; the latest thought, however, locates both in the second half of the first century ce. See appendix, and esp. Bowie (2002) 47–52 attacking the Hellenistic date for Ninus (but the positive case for the later date, it must be admitted, is more speculative). Braun (1934), (1938). 3 Overview at Humphrey (2000) 28–38. The hypothesis that Anthia and Habrocomes is epitomised (B¨urger (1892)) has been, in my view, effectively demolished by H¨agg (1966) and O’Sullivan (1995) 100–39. I offer some more thoughts on epitome theory and the romances in a forthcoming paper.

25

26

Part i Returning romance

and adapt traditional narrative patterns to articulate ideas of selfhood. Still, a quick sketch of the civic contexts in which Chariton’s and Xenophon’s romances appeared will help to explain some of the background. This is especially valuable for these two texts, since they are the most ‘civic’ of the romances, in that (particularly compared to the later romances) considerable emphasis is placed on beginning and ending in the polis, and the festivities and rituals that take place there. Given that urbanism is implicitly presented as the precondition for civilised life, it makes sense to consider what his own native city may have meant for each author. Chariton hailed from the Carian city of Aphrodisias, he tells us, and served as the secretary (hupographeus) of the orator Athenagoras (1.1.1). This identification, if it is not fictitious (‘Mr Favours, from the city of Aphrodite’ is a suspect name for an erotic romancer), is extremely interesting, since Aphrodisias occupied a unique position in Roman political history. As archaeology has demonstrated so dramatically,5 the city grew rapidly from a village to a major civic centre in the imperial period, thanks largely to the patronage of the emperors Augustus and Tiberius. It helped enormously, no doubt, that Aphrodisias had been uniquely loyal to Rome in general and the Julio-Claudians in particular throughout the turbulence of the first century bce.6 An inscription dated to 88 bce records that ‘our whole People, together with our wives and children and all our property [?] is ready [?] to risk all for Quintus and the Roman cause . . . without the rule of the Romans we do not even choose to live’.7 Inscriptions are not ingenuous expressions of sentiment, but they are contracts with fate. In this case, the decision was inspired. The city was granted free status by Octavian in 39 bce, perhaps renewing an earlier gift by Sulla. The Aphrodisians had declared their political hand early on, and reaped the benefits. Their theatre was dedicated by a freedman of Octavian. An extensive portico was built for Tiberius, and baths for Hadrian. Most spectacular of all was the Sebasteion, the centre of the cult of the living emperor, with its now-famous relief work allegorically depicting imperial victories over Armenians, Jews, Britons, Arabs and others, and the emperor’s rule over earth and sea.8 Aphrodisias was thus a monument to the regenerative power of benevolent imperial rule. 5

6

7

Up-to-date information and bibliography can be found at http://www.nyu.edu/projects/aphrodisias/ home.ti.htm. The inscriptions are usefully, and exceptionally well, catalogued at http://insaph.kcl. ac.uk/index.html. On Chariton and Aphrodisias, see now Tilg (2010) 24–36. For Aphrodisias’ support of Rome during the Mithridatic wars see App. Bell. civ. 1.97.455, with Reynolds (1982) d2–3 and p. 4; for loyalty to Caesar despite the invasion of Labienus, see Reynolds (1982) 5, 31. Reynolds (1982) d2 11–14. 8 Smith (1988).

First romances

27

This alignment of regional, panhellenic and imperial interests was embodied in the figure of Aphrodite herself.9 Aphrodisian representations of Aphrodite weave together three cultural strands. The first is a primeval Anatolian deity, or at least such a deity as imagined from the vantage of Roman times: the staid, unerotic cult figure with her distinctive ependyt¯es (or stiff tunic), recalling other Anatolian types such as the Artemis of Ephesus. The second is the familiar Greek goddess; the famous sculptural ‘school of Aphrodisias’ was particularly keen on Hellenistic depictions of the goddess. The third is the mother of Aeneas10 and progenitor of the Julio-Claudian dynasty. A statue group in the Sebasteion, for example, linked together Aphrodite (titled ‘the fore-mother (prom¯et¯or) of the divine Augusti’), Aeneas and ‘a wide-ranging selection of Julio-Claudian princes and princesses’.11 As so often in the imperial Greek world, cult served as the symbolic mediator between regional and panmediterranean interests. Aphrodite is also the central deity in Callirhoe. The egregiously beautiful protagonist is compared to or mistaken for her throughout.12 Her close connection with the goddess begins with a reference in the first chapter to her beauty, which is ‘not a human’s but a goddess’s; and not a Nereid’s or a mountain nymph’s, but of Aphrodite herself’,13 and ends in the prayer to the goddess that closes the narrative (8.8.15–16).14 Heavily influenced by the naturalistic conventions of historiography,15 Chariton portrays a world that is in general relatively free from divine intervention: aside from Aphrodite (and Eros), there are only passing references to Olympian deities.16 This is a text so light on deity that even Reinhold Merkelbach had to exclude it from his ‘mystery text’ theory of the Greek romance.17 Aphrodite’s central role is thus all the more significant. Throughout, she 9 10 11 12

13 14

15

16 17

For a comprehensive overview, see Brady (2007). A connection already made in the late republic: Reynolds (1982) 3–5. Smith (2006) 44. The base featuring the inscription is Aphrodisias inv. 82–117 in Smith (2006); cf. Reynolds (1986) 111. Callirhoe compared to or mistaken for Aphrodite: 1.1.2, 1.14.1, 2.2.6, 2.3.6, 2.5.7, 3.2.14, 5.9.1. Comprehensive discussion of Chariton’s Aphrodite at Alperowitz (1992) 41–57; useful remarks on Callirhoe as epiphany of Aphrodite also at H¨agg (2002) 52–5. See also Edwards (1991) 191–200; (1994); (1996) 20–2 (with index s.v. Chariton) for an Aphrodisian reading of Chariton’s Aphrodite. [§n . . . t¼ k†llov] oÉk ˆnqrÛpinon ˆll‡ qe±on, oÉd• Nhr¹dov £ NÅmfhv tän ½reiän ˆllì aÉtv %frod©thv, 1.1.2. The circularity is reinforced by the cross-reference to her initial prayer at 1.1.7–8: cf. esp. to±v posª pros”pese kaª katafiloÓsa (1.1.8) ∼ labom”nh . . . aÉtv tän p»dwn . . . katafiloÓsa (8.8.15). Foot-grabbing also appears at 2.2.7. Bartsch (1934), Zimmermann (1961), Hunter (1994), Alvares (1997), Manuwald (2000) 102–6. Morgan (1993) argues that the romances generally recur to historiography, to corroborate their craving for realism. As Weißenberger (1997), in particular, emphasises. Merkelbach (1962); a similar exemption is made by Merkelbach’s student Remi Petri (1963).

28

Part i Returning romance

(alone of gods) is prayed to in the expectation that she will influence events.18 Elsewhere she is reproached for her actions.19 On rare occasions, she is directly credited with ‘governing’ (politeuesthai) the plot, by arranging marriages.20 At the beginning of the final book, in a passage to which we shall return more than once, she is said to have overruled the plans of Tukh¯e (here in her guise as ‘Fortune’ or ‘Chance’ rather than ‘Providence’) to engineer another ‘gloomy’ (skuthr¯opon) event, since she ‘was now beginning to be reconciled (diallattomai) with [Chaereas], having previously been furiously angry (orgistheisa) with him on account of his inappropriate jealousy’.21 The spending of divine wrath – an epic motif – betokens the closure of the romance, and the harmonious restitution of order. More than this, Aphrodite reassumes her benign and providential role in the ordering of human affairs. In general terms, then, Callirhoe expresses an ultimate confidence in the tutelary beneficence of the established order, as embodied in the figure of Aphrodite, who mediates between local polis, transpolitical networks and cosmic order (the role she played, indeed, in Aphrodisian cult and architecture).22 She embodies what Edwards has called ‘the web of power’.23 Ephesus, whence Xenophon (whose modern surname ‘of Ephesus’ serves to distinguish him from his better-known Athenian namesake),24 was by contrast one of the oldest Ionian cities.25 Its position, on both the coast and the Maeander, lent it an enduring importance strategically and in terms of trade. Mythically, it laid claim to foundation by Androclus, the son of the legendary Athenian king Codrus; in fact, the city’s history had been largely shaped by the sixth-century Lydian king Croesus (who was behind the 18 20

21 22

23 24

25

1.1.7, 2.2.7–8, 3.2.12–13, 3.8.7–9, 6.2.4, 7.5.2–5, 8.2.8. 8.4.10, 8.8.15–16. 19 3.10.6, 5.10.1, 7.5.1–5. 2.2.8 (Šllon –politeÅeto g†mon); 5.1.1 (politeusam”nhv . . . t¼n g†mon). Eros, Aphrodite’s son (the relationship is stressed at 2.2.8), is also given this kind of match-making role (1.1.4, 1.1.6 (politeusam”nou), 1.1.12, 2.4.5; cf. also 4.7.5; 6.4.3–4). ¢dh . . . aÉtäi dihll†teto, pr»teron ½rgisqe±sa calepäv di‡ tŸn Škairon zhlotup©an, 8.1.3. See Edwards (1991), and esp. (1994) on Aphrodite as a figure for the ‘web of power’ in which Aphrodisias was implicated. Aphrodisians seem to have had a sustained interest in providence, to judge by both the cosmic imagery of the Sebasteion and the tract On fate by the Severan philosopher Alexander: Sharples (1983). Edwards (1994). The ‘surname’ has no ancient authority, appearing first at Suda X 50 (where of course the need to differentiate between homonyms is paramount). The Ephesian origin has sometimes been doubted – e.g. by Griffiths (1978), claiming that ‘our author is most at home in Lower Egypt, and especially Alexandria’ (426) – but the evidence is so slender (and the possible explanations for distortions of or accuracy in any particular local knowledge so numerous) that I have opted to stick, tentatively, with the traditional identification. Bammer (1988) offers an accessible sketch, written by a major expert; see also (1984), narrowly on the Artemision. Knibbe (1998) 59–235 gives a narrative survey; see also the essays in Koester (1995). The Roman context is summarised by Rogers (1991) 2–16.

First romances

29

famous cult of Artemis Ephesia, incorporating both Greek and Anatolian elements), and by Lysimachus in the early third century, whose replanning of the city allowed it to develop into the spectacular environment still visible today. Unlike Aphrodisias, Ephesus had a history of non-compliance: it seceded from the Delian League in c. 412 bce, refused an offer of cashinjection from Alexander the Great, and – most importantly for our purposes – sided with Mithridates against Rome in the first century bce, despite forming part of Attalus III’s bequest to the Romans in 133. The citizens tore down Roman statues, it is said (App. Mithr. 21), and massacred even those who had sought sanctuary with the goddess (ibid. 23). For this the Ephesians were punished by a vengeful Sulla, who, not content with fining the city and removing its freedom, reserved a most grievous treatment for this city, one that our sources do not relay (ibid. 61). Even so, by Xenophon’s time the city’s natural advantages, its status as an assize centre and the reorganisation of the provinces under the empire (which limited the opportunities for corruption) meant that it had regained its position of primacy in Asia – particularly in matters of imperial cult, in respect of which the Flavian emperors awarded it the prestigious title of neokoros (‘minister’), apparently the first time that the term had been applied to a city in this way.26 Ephesus is presented in Anthia and Habrocomes as the centre of the civilised world, together with Rhodes (another city that benefited from Roman favour, thanks to its loyalty in the Mithridatic wars). The patron deities of the two cities, Artemis and Apollo-Helios, steer the narrative (as we shall see in greater detail below). Unlike Chariton, Xenophon (if the biographical tradition is accurate) set his romance in his own city. According to the Suda (X 50), he wrote another work On the city of Ephesus; this was perhaps periegetical in nature.27 In line with the general deprecation of our author, many scholars have uncharitably downplayed Xenophon’s awareness of epichoric features.28 It is true enough that his Artemis is the familiar panhellenic goddess of hunting rather than the protuberant Anatolian hybrid, but this does not mean that Xenophon’s image is unEphesian. It is misleading to allow our idea of the Ephesian cult to be exclusively dominated by the latter image. We do, in fact, have a snapshot 26 27

28

Friesen (1993), esp. 50–74. Xenofän ìEf”siov, ¬storik»v. ìEfesiak†á ›sti d• –rwtik‡ bibl©a © perª ìAbrok»mou kaª ìAnq©avá kaª perª tv p»lewv ìEfes©wná kaª Šlla. It is occasionally speculated that the phrase ‘and concerning the city of Ephesus’ may be a descriptive gloss also referring to the romance (e.g. Kytzler (2003) 346), but this would be an anomalous way for the Suda to refer to a romance; and moreover kaª Šlla suggests an itemised list of works. G¨artner (1967) 2058–9; Griffiths (1978) 426.

30

Part i Returning romance

of at least one facet of the Ephesian cult of Artemis in the early Roman period, in the form of the celebrated inscription of Gaius Vibius Salutaris, set up in 104 ce (so approximately contemporary with Xenophon, on one dating), which stipulates (among other things) a procession of statues to the famous temple.29 There are many differences between this procession and the festival near the start of Anthia and Habrocomes (1.2) – in particular, the former mixes adulation of the emperor (together with the senate and other Romans), a synoptic history of the polis, and self-promotion on its founder’s part, whereas the latter is presented as a traditional civic festival30 – but, crucially for our purposes, the inscription represents the goddess, as Xenophon does, in purely panhellenic guise. Of course, we have no way of knowing what the Artemis statues mentioned in the Salutaris decree looked like – perhaps they did depict the goddess in her epichoric form – but the inscription avoids the opportunity to regionalise her presentation, and arguably even actively panhellenises her. The only iconographic markers mentioned in the decree itself are stags (elaphoi, l.159) and torches (lampades, ll. 164–5, 168, 173 [restored], 186–7, 194 [restored]),31 elements that are common to the conventional iconographic repertoire. Both elements are connected with the goddess in Xenophon.32 In other words, Xenophon’s Hellenised version of Artemis may well reflect not the author’s ignorance of genuine Ephesian cult, but precisely the opposite: a tendency among contemporary Ephesians to downplay non-Greek elements. Chariton and Xenophon, then, both originated in Asia Minor, and indeed their romances emerge from the provincial (and also imperial) polis culture that is so well attested to epigraphically. Asia Minor, indeed, seems particularly important for the early romance. The fragmentary romance that modern scholars call Ninus may well have originated from Aphrodisias, like Callirhoe: an early name for the city had been Ninoe, and two panels of the basilica reliefs portray, respectively, Ninus and Semiramis in Hellenised garb.33 (It is possible, although wholly unprovable, that Chariton even 29 30 31 32

33

IE 1a 27; Rogers (1991) controversially argues that it embodies the people’s anxiety over the identity crisis generated by Roman intervention. Rogers (1991) 81: ‘no ritual acts took place during the performance of Salutaris’ procession. The procession did not form one of the constituent parts of the major religious festivals in the city.’ Rogers (1991) 111 speculatively connects these with chthonic rites and the cult of Cybele, but they are too common a motif in Greek religion to convey these associations unequivocally. Anthia, who is dressed as the huntress-goddess, wears a fawn-skin tunic (1.2.6); some in the procession carry torches (daides, 1.2.4). A further point of comparison is the significant role given in Salutaris’ procession to the ephebes (l. 211 [restored]), who are, along with the parthenoi, the focus of Xenophon’s procession. On the significance of the Ninus narrative to Aphrodisias, see Yildirim (2004).

First romances

31

wrote Ninus: the dates and styles seem to fit).34 Another ‘historical romance’ seemingly from the same period, and apparently stylistically similar to Callirhoe, is the fragmentary Metiochus and Parthenope, set on Samos (just off the Ionian coast).35 The Ionian coast also provided the settings for other erotic narratives, now lost (and hence undatable): the Cypriaca of Xenophon of Cyprus (Suda X 51: probably a scurrilous local history); and Philip of Amphipolis’ Rhodiaka, an ‘absolutely disgraceful’ work, a pious Byzantine encyclopedia tells us (Suda F 351). What does this localisation in Asia Minor tell us? The first point is that Ionia had suffered terribly at Roman hands in the late republic, as war and the depredations of governors took their toll. By the mid-first century ce, to which period our earliest romances seem to date, however, the region as a whole, and in particular favoured cities like Ephesus and Aphrodisias, had benefited from provincial centralisation, the prosecution of exploitative officials (e.g. Tac. Ann. 3.66–9, 4.15) and investment. The first-century romances, as we shall see, celebrate the capacity of Greek cities for post-crisis rejuvenation under benevolent authority. In this respect, they mimic Julio-Claudian imperial rhetoric, which also emphasised the theme of renewal at the cultural, moral and natural-cosmic levels, often imaged in terms of fecundity and legitimate reproduction.36 The texts are not, however, simply allegorical celebrations of the benevolence of foreign empire. The happy ending is closely tied to return to the specifically Greek polis; foreigners and foreign lands, by contrast, represent hostility and threat. Xenophon and Chariton are deeply ambivalent towards power, as we shall see in more detail below. One recurring typescene involves a lusty barbarian wheedling and making threats towards one of the lovers if she or he does not grant sexual favours; the latter nonetheless remains true to his or her beloved.37 There are close links between these scenes and martyr accounts – including, as well as the Jewish and Christian texts, the fragmentary Gentile ‘acts of the pagan martyrs’ – where virtuous individuals stand up to thuggish despots, particularly Romans.38 As so often in the literature of the Roman era (and, indeed, in Hellenistic Jewish narrative),39 aggressive power relations, particularly in the sphere of 34 35 36 38 39

Bowie (2002) 49–51. Edited, together with a text and translation of the Coptic and Persian Nachleben, in HU. Zanker (1988) 101–66. 37 Char. 6.5; Xen. Eph. 2.3–6. For the similarities with the Acta Alexandrinorum, see APM 252–8; and, on the similarities with Christian martyr acts, Perkins (1995) 41–76. Braun (1938) 44–102 on Hellenistic Jewish texts, principally the Greek Testaments of Joseph and Reuben (though the Hellenistic dating of these is now the matter of some controversy), and more

32

Part i Returning romance

sexuality and interpersonal ethics, serve as allegories for the relationship between imperial power and subject. In fact, all forms of hierarchical power are treated with caution in the first-century romances: the safety of the individual is only guaranteed in the context of the Greek city, where festivals and assemblies signal the enfranchisement of the entire populace, under the tutelage of a benign oligarchy. Not until Heliodorus in the fourth century do we encounter in the romances any intimation that happiness is reconcilable with monarchical rule. Ultimately, in these earlier texts it is not emperors and kings who guarantee the social order of Greek polities, but gods.40 Gods, indeed, are politically ambiguous figures: they can be taken either as figurative of the emperor’s power, or contrariwise as emblems of what mortal power is not, viz. absolute, uncontested and permanent. (This ambiguity of divinity reflects one of the roles of cult in Greek provincial poleis, as we saw above, namely simultaneously to emblematise distinctive regional sacrality and transmediterranean interconnectivity.) It is thus misguided to ask straightforwardly whether the romances promote or attack Roman hegemony. Like much early imperial Greek cultural production, they avoid talking directly about Rome, preferring to allegorise power by translating it into other realms (particularly those of ethics and religion), and then treating it in a complex, multifaceted way. social crisis What is certainly the case, however, is that the first-century romances construct the health of the polis as the essential precondition for the happiness of the individual. The commonly expressed view that the romance is constitutively post-political – ‘a social and personal myth, of the private individual isolated and insecure in a world too big for him’41 – could not, thus, be further from the truth. The polis is at the heart of Chariton’s and Xenophon’s conception of civilised life (a fact that emerges even more strikingly through the contrast with the later romances of Longus

40

41

generally Shaw (1986). Interestingly, a similar scene is found at Joseph and Aseneth 23, where the sexual aggression of Pharaoh’s son defines by contrast the virtues (vigour and wisdom respectively) of the Jews Simeon and Levi. On the issues around the dating of Joseph and Aseneth see above, n. 3. Perhaps reflecting the central role of the gods in dispute resolution, in contemporary cities: Hellenistic and imperial inscriptions from Asia Minor represent the gods as arbiters of social justice (Chaniotis (2004)). Reardon (1991) 28–9; see also above, pp. 8–9.

First romances

33

and Achilles).42 The egregiously beautiful youths upon whom the narrative centres are iconic for the city. Callirhoe is ‘a marvel of a girl, the cult statue of all Sicily’;43 Chaereas has ‘a youthful beauty surpassing all others’, and ‘gleams like a star’.44 The language suggests the effulgence of a Pindaric victor, in the eyes of his compatriots. In return for the lustre bestowed, the people treat the protagonists with a superhuman reverence. Xenophon’s Habrocomes ‘was cultivated by all the Ephesians, and even all the inhabitants of Asia . . . they treated the young man like a god, and some even prostrated themselves and prayed to him when they saw him’.45 Habrocomes and Anthia appear at the head of the troupes of young girls and boys in the festival processions (1.2), and again the people take a central interest: ‘What a marriage it would be between Habrocomes and Anthia!’ they shout.46 In Chariton, it is the people, again, who agitate for the marriage between Chaereas and Callirhoe, a marriage that the ‘patriotic’ (philopatris, 1.1.12) Hermocrates, Callirhoe’s father, feels unable to refuse; consequently, ‘the Syracusans celebrated that day with more joy than the day of their victory over the Athenians’.47 The young lovers’ effulgence at one level separates them from the masses. As so often in aristocratic Greek culture, physical beauty accompanies (and, ideologically speaking, serves to naturalise) social distinction. Yet it would be a misunderstanding of aristocratic ideology to suggest that the elite are somehow isolated from the rest of the city. Rather, what distinguishes them is their ability to distil and embody the city’s most sublime qualities. In this

42

43 44

45

46 47

Cooper (1996) 20–44 and Swain (1996) 101–31 both emphasise the role of civic values, but without distinguishing the registers of different romances; Morales (2008) 42–3 suggests a more pluralist approach. qaumast»n ti crma parq”non kaª Šgalma tv Âlhv Sikel©av, Char. 1.1.1. meir‰kion eÎmorfon p‰ntwn Ëper”con, Char. 1.1.3; st©lbwn ãsper ˆstžr, Char. 1.1.5. This stellar imagery, which goes back to Homer (Il 5.5, 22.26), is particularly common in lyric depictions of egregious individuals: discussion and references at Whitmarsh (2004b) 388–90. It is also found in Joseph and Aseneth (2.6). §n d• perispoÅdastov Œpasin ì Efes©oiv, ˆll‡ kaª to±v tŸn Šllhn %s©an o­kousi . . . prose±con d• Þv qeäi täi meirak©wi, kaª e«sin ¢dh tin•v o° kaª prosekÅnhsan kaª proshÅxanto ­dontev, Xen. Eph. 1.1.3. o³ov ‹n g†mov g”noito &brok»mou kaª %nq©av, Xen. Eph. 1.2.9. ¤dion taÅthn tŸn ¡m”ran ¢gagon o¬ Surak»sioi tv tän –pinik©wn, Xen. Eph. 1.1.13. Alvares (1997) 619 notes also the pandemic support for a mission to rescue Callirhoe, once it is realised that she is still alive: ‘only in Chaereas and Callirhoe does the entire state strive to reunite the lovers, as if it were a matter of the highest political import’; ‘this corporate mission to regain Callirhoe’, he also observes, ‘recalls the mythical panhellenic effort to recover Helen’. The analogous episode in Xenophon has ‘the entire people’ praying and sacrificing for Anthia and Habrocomes prior to their departure, (1.10.5), and the ‘entire mass of Ephesians’ seeing them off (1.10.6; plqov is Hemsterhuis’ supplement, but a plausible one: a neuter noun is clearly needed here).

34

Part i Returning romance

normative model, there is no class friction: the happiness of the aristocracy is celebrated with just as much joy by the people. Civic harmony, however, is put at risk by the onset of desire, which creates the narrative energy that stimulates the plot. As Edward Said has argued, the beginnings of literary works establish contracts with the reader, marking the state against which the remainder of the narrative defines itself, by more or less deviant forms of repetition.48 Chariton and Xenophon exemplify this in a distinctive, beguilingly simple way. The beginning defines a state of social and psychic stability that is compromised by the onset of love. The subsequent narrative becomes an attempt to recapture that state; only at the end is it recaptured, but (as we shall see in the course of this chapter) in a way that suggests renewal as well as return. Chariton (after his one-sentence prologue) and Xenophon show a remarkable degree of congruence in their handling of their beginnings. Each describes an egregiously beautiful young man and woman, who fall in love at a festival – despite powerful reasons why they should not (in Chariton because their parents are political enemies, in Xenophon because Habrocomes shows a Hippolytus-like distaste for sexuality). The differences will emerge presently, but the similarities are so close that they can be tabulated (see Table 1 below). I leave aside the unanswerable questions of which author follows which, or whether both follow a common source;49 what is more important to us here is that both describe a similar process. Each emphasises the disruptive power of desire on the community, a case of what the Russian critic Boris Tomashevksy calls ‘the exciting force’: ‘In order to get the story going, a dynamic motif destroys the initial peaceful situation.’50 The peaceful situation is expressed at the level of discourse by the use of the imperfect and present to describe them in the sections I have marked as A and C in Table 1.51 This ‘iterative’ mode of narration exists in narrative primarily to be interrupted by dynamic events.52 Sure enough, Eros intervenes, bringing about the transition from stasis to the dynamism of narrative: again, this latter state is marked discursively by the use of past tenses denoting change and rapid action (‘singulative action’).53 48 51 52 53

Said (1975). 49 See appendix. 50 Tomashevsky (1965) 72. The exception is –x”balen of Habrocomes at Xen. Eph. 1.1.5, where one would expect an imperfect (this may be a corruption: Hemsterhuis proposes –x”ballen). Genette (1980) 116–17: ‘iterative scenes are almost always functionally subordinate to singulative scenes, for which the iterative sections provide a sort of information frame’. In Chariton, the first aorists of the text (¾ d• ï Erwv zeÓgov «dion  q”lhse sull”xai . . . –zžthse d• toi»nde t¼n kair»n, Char. 1.1.3); Xenophon however, uses inceptive imperfects rather than aorists (though NB the aorist participles –xopl©sav . . . paribal»menov, Xen. Eph. 1.2.1).

First romances

35

Table 1: Similarities between the texts Chariton

Xenophon

A. Description of Callirhoe (1.1.1–2)

A. Description of Habrocomes (1.1.1–6)

B. Eros’ desire to create an unusual match (1.1.3)

B. Eros’ wrath at his contumely, and search for a plan (tekhn¯e, 1.1.6)

C. Description of Chaereas (1.1.3)

C & E. Local festival (epikh¯orios heort¯e) where lovers meet, and description of Anthia (1.2.1–9)

D. Eros’ pleasure in paradoxes, and search for an opportunity (kairos) (1.1.4)

D. ‘Such were the devices of Eros’ plan (tekhn¯e)’ (1.2.9)

E. Public festival (heort¯e d¯emotel¯es) of Aphrodite and subsequent meeting of lovers (1.1.4–6) F. Erotic sickness (1.1.7–10)

F. Erotic sickness (1.3–5)

G. Diagnosis (1.1.10), followed by cure = marriage (1.1.11–16)

G. Diagnosis (by oracle and interpretation: 1.6–7) followed by cure = marriage (1.8)

Eros is an urgent and vital, but perverse and disruptive, force. In Chariton he is said to want to make ‘his own kind of (idion) match’,54 the phrasing indicating that he has personal ambition rather than the collective good at heart. For this reason, we are told, he brings together the children of two political rivals; for, the narrator comments apophthegmatically, ‘he is competitive (philoneikos), and rejoices in successes against the odds’ (1.1.4).55 In Xenophon, he is also described as ‘competitive’ (1.2.1). This, in Xenophon, leads him to ‘rage’ (m¯eni¯ai, 1.2.1) against Habrocomes, an anger that no doubt alludes to the comparable narrative-inceptive role as the ‘rage’ (m¯enis) of Achilles, thematically announced in the first line of Homer’s Iliad. Eros thus embodies the qualities of rivalrous individualism that the Greeks considered so destructive of civic unity.56 The competitiveness 54

55

56

zeÓgov «dion (context cited in previous note). I have offered the conventional interpretation of this phrase (which, I believe, fits better with the emphasis upon Eros’ idiosyncratic behaviour); but «dion might also be taken as ‘local’ (i.e. in contrast to the suitors from Italy and further afield just mentioned). For similar characterisations of Eros, see 2.4.5 (–filone©kei), 6.4.5 (fil»nikov). TÅch shares this attribute: see 2.8.3, 5.1.4, 6.8.1. I have retained MS readings contra Reardon who prefers to standardise to fil»nik-, but the palaeographical and semantic conflation of the two (homophonic) terms is notorious (good discussion at Duff (1999) 83). Konstan and Rutter eds (2003).

36

Part i Returning romance

that characterises him metonymically figures his effects upon society. In Chariton, notably, the wedding of Chaereas and Callirhoe is compared to that of Peleus and Thetis, where (the near-namesake of Eros) Eris, the personification of strife, began the chain of events leading to the Trojan War (1.1.16): the frustrated suitors gather and plot (1.2), and deliberately engineer in Chaereas the jealousy (z¯elotupia, 1.2.5–6) that causes him to attack Callirhoe (and indeed energises numerous later scenes in the narrative).57 Eros is thus at one level a god, an external force who (like Aphrodite in Euripides’ Hippolytus) afflicts individuals against their will; at another, however, he serves as a figure for the psychic turbulence – the arrogance, competitiveness, frustration and identity confusion – that comes with adolescence.58 Eros thus also plays a metanarrative role: he represents the principle of inventiveness and change necessary for a story to progress. He is a strategic plotter: in both texts he is said to ‘search’ (z¯etein), respectively for an ‘opportunity’ (kairos, Char. 1.1.4) and a ‘plan’ (tekhn¯e, Xen. Eph. 1.2.1, 1.2.9). Discussing the modern novel, Peter Brooks comments on the regularity with which desire figures the inception of narrative: ‘Desire is always there at the start of a narrative, often in a state of initial arousal, often having reached a state of intensity such that movement must be created, action undertaken, change begun.’59 The first-century romance alchemically transforms highlevel theory into narrative praxis (a phenomenon that we shall see again): Eros figures, and also embodies, the magical process whereby narrative energy is created apparently out of nothing, and stillness suddenly becomes chaos. From now on until closure is reached, life for the protagonists is to be experienced in the liminal phase, with its hyperactive, staccato sequence of ‘congealed “suddenlys”’.60 At the same time, however, Eros also embodies the intention of ultimate closure, of legitimate consummation and the final spending of that sexual-narrative energy. Desire exists, at least for the protagonists of the romance, to be ultimately satisfied. In the first-century 57

58

59

For zhlotup©a, see also Char. 3.9.4, 4.4.9, 5.1.1, 5.9.9, 6.6.5, 6.6.7, 6.7.11, 8.1.3, 8.1.15, 8.4.4, 8.5.15, 8.7.6; Helms (1966) 32–4. Fantham (1986) posits a fundamental link between zhlotup©a and violence between the sexes, beginning from the pun at Arist. Plut. 1014–16 (–tupt»mhn . . . zhlotÅpov); she also discusses the stock comic motif of the violently jealous lover. But violence towards pregnant women (not that Chaereas is aware of his wife’s state) is specifically characteristic of tyrannical figures: see Ameling (1986). Chariton is particularly fond of this allegorical approach, whereby the subjective emotions stimulated in humans by er¯os are attributed to the god himself: ‘Eros is by nature optimistic (fÅsei . . . eÎelp©v –stin ¾ ›rwv, 2.6.4); ‘the love of ornamentation is a characteristic of Eros’ (›sti . . . ­dion ›rwtov fil»kosmon, 6.4.3). Ancient orthography, of course, did not distinguish proper name from abstraction. Brooks (1984) 38. 60 Bakhtin (1981) 102.

First romances

37

romance, that process will coincide with the healing of differences within the polis, the restoration of social concord and psychic wholeness – and also the reader’s consummation of her or his desire for narrative telos. In both Chariton and Xenophon, falling in love catalyses a crisis in the community. The love-sickness experienced by the lovers is replicated at the civic level, as negative, destructive emotions surface, threatening the cohesion of the social group. This play between social/psychic wholeness and fragmentation is articulated by ritual practice, particularly in Xenophon. I mean ‘articulated’ in a double sense. Firstly, ritual is the primary ‘language’ through which these movements are expressed. Secondly, and relatedly, rituals mark the ‘joints’ (articuli) of the narrative, the junctures where societies either reconstitute themselves or decompose into discord.61 In Chariton, it is at a ‘public festival’ (heort¯e d¯emotel¯es, 1.1.4), in Xenophon at a ‘local festival’ (epikh¯orios heort¯e, 1.2.2), that the lovers meet. This idea that festivals are occasions of sexual possibilities is, of course, a staple of erotic literature,62 and presumably has a certain basis in the reality of ancient Mediterranean life, where such occasions offered a rare opportunity for women and men to commingle. This clich´e is reanimated in the romance, however, as the festival metamorphoses from a topical narrative device into a feature in its own right – again, particularly in Xenophon, where it receives an uncharacteristically lavish ecphrasis (1.2.2–9). This description mobilises all the civic themes of the opening section of the romance, bringing the relationship between individual and community into sharp focus. The central dynamic, for Xenophon, is the dialogue between the masses of spectators (both local and foreign: 1.2.3)63 and the youthful procession. The spectators figure a community united in the act of viewing. The objects of their gaze, meanwhile, ritually perform the harmony of the occasion. Maidens and ephebes are separated; they process ‘in file’ (kata stikhon, 1.2.4). Marriage is the expected outcome, since ‘it was the custom at that festival for grooms to be chosen for maidens, and wives for ephebes’.64 Marriage figures social concord, as so often in this period:65 the coming together of women and men, maidens and ephebes at the culminating sacrifice (1.3.1) emblematises 61 62

63 64 65

Lalanne (2006) 118–22. Rohde (1914) 155; Trenkner (1958) 110. See e.g. Lys. 1.20 (perhaps a source for Chariton: Porter (2003)); Lycophr. 102–9; Plaut. Cist. 89–93 (perhaps ∼ Men. fr. 382 K¨orte); Call. Aet. 67.5–8, 80+2 Pfeiffer; Parth. Am. 32.2; Jos. Ant. Jud. 2.45 (dhmoteloÓv . . . —ortv, Chariton’s phrase: Whitmarsh (2007b) 88). Though in fact subsequently only Ephesians are mentioned: 1.2.7. kaª g‡r ›qov §n –ke©nhi ti panhgÅrei kaª numf©ouv ta±v parq”noiv eËr©skesqai kaª guna±kav to±v –fžboiv, Xen. Eph. 1.2.3. On the shared vocabulary of political and marital concord in this imperial era see Veyne (1978).

38

Part i Returning romance

the complementarity of opposed elements. The community, meanwhile, invests its aspirations in the projected marriage of the two most beauteous youths: we have already considered their cry, ‘what a marriage it would be between Habrocomes and Anthia!’ (1.2.9). There is, however, an ominous underlying theme. Even during the happy festival, we know that Eros has promised discord: the description is introduced by way of explicating the ‘device’ (tekhn¯e) that this ‘strife-loving’ (philoneikos) god sought to use against Habrocomes. This discord is figured in Eros’ anger (he ‘raged’, m¯eniai: 1.2.1) and military apparel (‘he armed himself, and equipping himself with the entire force of his erotic drugs, began his campaign against Habrocomes’).66 This aggressive, militarist iconography leaks into the description of the parade: horses, dogs and hunting equipment are paraded (we read), are a mixture of the ‘warlike’ (polemika) and the ‘peaceful’ (eir¯enika).67 The community’s investment in the pair, moreover, is immediately rewarded not with happy marriage but with agonising sickness, which brings each of them close to death. Even when the two lovers are married, the discordant undertones continue. The marriage itself reprises the theme of pandemic happiness, and harmony itself engulfs the entire community (‘the city was full of merrymakers’; ‘all-night parties were held and many sacrifices were offered to the goddess’).68 Once again, though, the festal occasion is heavily overshadowed: by the oracle’s dark intimations; by the foreknowledge that their parents have decided to send them abroad for a certain time (1.7.2); and by the iconography of the coverlet on the marital bed, which presents the Erotes waiting on Aphrodite on one side, but the adultery of Aphrodite with the warlike Ares on the other (1.8.2–3). In Chariton, characteristically, these themes of harmony and conflict emerge more naturalistically, motivated by personal and interpersonal psychology. Callirhoe goes to a festival of Aphrodite, but rather than meeting Chaereas there (as the serendipitous episode in Xenophon would lead us to predict), she happens to bump into Chaereas (who is coming back from the gym) on a street corner afterwards (1.1.5). The festival is incidental to the action, and not described in any detail; it exists sous rature, as though Chariton were reminding us of the clich´es adopted by lesser romancers. 66 67 68

–xopl©sav . . . —aut¼n kaª psan dÅnamin –rwtikän farm†kwn peribal»menov –str†teuen –fì &brok»mhn, Xen. Eph. 1.2.1. †ædrª† polemik†, t‡ d• ple±sta e«rhnik†, Xen. Eph. 1.2.4. Despite the irresoluble corruption in the initial part of the sentence, the opposition seems secure. mestŸ . . . ¡ p»liv §n tän eÉwcoum”nwn, 1.7.3; pannuc©dev ¢gonto kaª ¬ere±a poll‡ –qÅeto ti qeäi, 1.8.1.

First romances

39

The tensions, instead, are played out at the political level, and largely in civic space.69 Chaereas’ and Callirhoe’s parents have a ‘political hostility’ (politikos phthonos), which is specifically activated by the children’s passion for each other (1.1.3). Hermocrates, Callirhoe’s father, only assents to the marriage because a ‘legal assembly’ is held, and the populace insists upon the wedding; the ‘patriotic’ (philopatris) Hermocrates cannot refuse the will of the people and Eros ‘the demagogue’ (1.1.11–12). Sexual awakening in young women and men thus forces them to become pawns, and to an extent players, in the political game of dynastic matchmaking. This game is a complex one, accommodating the partially conflicting will of parents and family, the community at large, and of course the lovers themselves. The discordant notes are sounded not so much (as in Xenophon) by militaristic imagery at the abstract levels of divinity and ecphrastic description, as at the naturalistic level, through the representation of the vagaries of interpersonal conflict and their psychological consequences. The general joy at the marriage is offset by the grief (lup¯e ) and anger (org¯e, 1.2.1) experienced by the band of frustrated suitors (mn¯est¯eres, distantly evoking Penelope’s suitors in the Odyssey). It is these individuals who directly embody the imagery of militarism and strife that is more abstract in Xenophon.70 Moreover, whereas Xenophon’s Eros, a deity, seeks a ‘device’ (tekhn¯e, 1.1.6) to use against Habrocomes, it is Chariton’s human suitors who perform the analogous scheming function (tekhn¯e, 1.2.4; cf. 1.4.1). In Xenophon, the work of disruption and disordering is part of an inscrutable and (in realist terms) undermotivated metanarrative plan; in Chariton, that same work is done by human agents, whose reasons are motivated ‘realistically’.71 The beginnings of the first-century romances dramatise societies performing their collectivity through time-honoured ritual and political practice – at the same time as the failure of that collectivity, its capacity for self-destruction under pressure. The narrative of separation from the community can thus be read both as a narrative narrowly about named individuals, and more generally as an allegory for the psychic and social trauma 69 70

71

Connors (2008) 164–5; Morales (2008) 42–3. Military imagery: –stratol»gei . . . aÉtoÆv –pª t¼n kat‡ Cair”ou p»lemon, Char. 1.2.1; ceirotonžsat” me toÓ pr¼v Cair”an pol”mou strathg»n / –fopliä / sÅmmacon, Char. 1.2.5. They are also associated with athletic agonistics: ßsper –n to±v gumniko±v ˆgäsin, Char. 1.2.2; basil”wn ˆgwnisam”nwn aÉt¼v ˆkonitª t¼n st”ganon ¢rato, Char. 1.2.3 (a ‘naturalistic’ counterpart to Habrocomes’ figurative wrestling match with Eros, Xen. Eph. 1.4.1–5); qlon, Char. 1.2.4. Similarly, whereas Chariton’s intercontinental chase is motivated by a chain of naturalistic events, Xenophon’s lovers are packed off on their travel simply to ‘appease’ (paramuqžsasqai, Xen. Eph. 1.7.2, 1.10.3) an oracle that is itself undermotivated (all the parents wanted to know was why their children were pining away).

40

Part i Returning romance

that comes with burgeoning sexuality and the prospect of marriage: the complex networks of obligation and competition, the pain and confusion, the end of the unconditional bond between parents and children. beginning, middle, end How are these beginnings set against the rest of the narrative? As Aristotle famously observed, every story needs a beginning, a middle and an end (even if, as Jean-Luc Godard is said to have opined, ‘not necessarily in that order’).72 Although (of course) there are variations at the level of cultural practice, this schema, in its most fundamental form, encodes the stark reality of our existence: we are born, we live, we die. All forms of human cultural expression have been, at some level, attempts to manage this awful truth. If the story of birth, life and death is humanity’s primordial tragedy, its narratives of redemption and transcendence are no less important. To be human requires not just acceptance of one’s own mortality, but also faith in the survival of humanity beyond the individual’s death: the end of one generation is the beginning of another. Sexuality is our glimpse of immortality, our compensation for death. The story of sexual fulfilment is again tripartite – we desire, we pursue, we consummate – but its ending is joyous and celebratory. The story of human sexuality is also the story of human sociality, of how the frail transitoriness of the individual is offset by the permanence of the family and the community. It is also, at a more abstract level, the story of religion: of the benign rationality of the divine order, of the indestructibility of the soul, of reproduction as a metaphor for eternity. This second kind of tripartition underpins the romance. A girl and boy of marriageable age meet, fall in love, endure numerous obstacles, and then are joyously reunited in enduring matrimony at the end. In Chariton and Xenophon, the marriage occurs at the start, and is then reinstituted at the end; in the later romances, as we shall see in the following chapters, the marriage is shifted to the end. This narrative tripartition of the adventure romance is well-known, but has often been misinterpreted. The first point to make is that it is not a primitive, overschematic attempt to narrate psychological development. In an influential account, Mikhail Bakhtin argues that the disjunction between middle and end serves only the crude purpose of separating the two significant episodes located at the beginning 72

Arist. Poet. 1450b 25–7; Godard quoted in Sontag (1969) 157.

First romances

41

and the end, namely the conception and the satisfaction of desire. This ‘gap’, he continues: the pause, the hiatus that appears between these two strictly adjacent biographical moments and in which, as it were, the entire romance is constructed is not contained in the biographical time-sequence, it lies outside biographical time; it changes nothing in the life of the heroes, and introduces nothing into their life. It is, precisely, an extratemporal hiatus between two moments of biographical time.73

For Bakhtin, then, the ‘hiatus’ after the moment of infatuation – the middle of the narrative – serves purely to delay the moment of consummation: there is otherwise no meaningful development during this period, which ‘leaves no trace in the life of the heroes or their personalities’.74 Bakhtin’s analysis is in fact entirely negative, sensitive only to what the romances lack, viz. a cogent psychology. John Morgan has rightly observed that no ancient romancer ever composed a Bildungsroman – ‘Love is not developed qualitatively, nor are new insights into the self achieved’75 – but this kind of claim expresses precisely the problem behind the formulation of the question. Would we expect any ancient text to deal with psychology in this way? Antiquity is not to be berated for its failure to promote the enlightenment conception of the self as ‘subjective’: consciousness-centred, entirely internalised, independent of external influences.76 What the romances narrate is not the protagonists’ acquisition of mature selfhood – understood in terms of autonomous subjectivity – but the changing ways in which the individual can be understood as a social and ethical being in relation to communities: an objective-participant conception of the self, to use Christopher Gill’s phrase.77 Callirhoe and Anthia and Habrocomes are, as we have said, fundamentally about social life in one’s native polis, and the traumas that occur when one 73

74 75 76

77

Bakhtin (1981) 87–90, at 89–90. For an assessment of Bakhtin’s value to criticism on the ancient romance, see especially Branham (2002), and (2005); and for a more critical account, Whitmarsh (2005d). Bakhtin (1981) 90. Morgan (1996), at 188. For the claimed lack of character development see also Konstan (1994) 45–6. Laplace (1991) and (2007), however, argues for development in Achilles Tatius. The bibliography is enormous. Especially useful philosophical studies are Long (1991), Gill (1996), and esp. (2006) xiii-xiv: ‘what is innovative and distinctive in Hellenistic-Roman thought about selfhood is not, as is sometimes claimed, a shift towards a heightened interest in subjectivity’. The contributors to Pelling ed. (1990) approach the problem from a variety of literary and philosophical perspectives. Duff (1999) 13: ‘Ancient conceptions of character were . . . less centred on the private, inner world of the individual’; they were concerned ‘more with actions, and their evaluation’. For the relatively limited role of ‘character change’ in Plutarch, see Gill (1983), Swain (1989), Pelling (1990), and further below pp. 214–20. Gill (1996), (2006).

42

Part i Returning romance

is cut off from that. Xenophon even seems to have granted Ephesus equal billing in the full version of the title, namely The Ephesian affairs of Anthia and Habrocomes.78 Both are built around a mythical structure of centre and periphery:79 the protagonists begin at home, travel abroad into marginal space before returning home at the end. In the simplest terms, the setting of the majority of the suffering narrative in ‘barbarian’ spaces is underpinned by a self-other model, reinforcing the cultural and ethical polarity of Greek and barbarian so frequently articulated by the characters.80 The ending thus represents something more than mere affirmation: it is a powerfully symbolic, redemptory celebration of the Greek polis as the figurative and literal centre of a world that can ultimately be recognised to be overseen by deities with the interests of civilised, aristocratic Greeks at heart. The centre–periphery structure creates powerful resonances. Greek culture contains numerous accounts of young women and men at the critical period of adolescence, separated from their society for a period and then reintegrated as adults. Van Gennep’s hugely influential study of 1908, Les rites de passage, offers a template for the analysis of initiation in terms of three stages: pre-liminal (rites of separation), liminal (rites of transition, experienced in a state of separation), and post-liminal (rites of reincorporation into the world).81 The model of ritual marginalisation embodied in myths such as those of Callisto, Io, Orestes, Telemachus, Jason and various young hunters (Odysseus, Meleager, Actaeon, Hippolytus) seems to have been mirrored (although the extent of this remains controversial) in civic institutions such as the ephebeia attested in various cities, the arkteia at Brauron, and the Spartan krypteia.82 The bulk of scholarly work has

78 79

80 81

82

Whitmarsh (2005b) 598–9. For the traditional division of space into centre and periphery, see Vidal-Naquet (1986a) 138–9. It is impossible to determine what structure the fragmentary romances adopted: the evidence is most plentiful for Metiochus and Parthenope, but still ambiguous (HU 247–50). See esp. Bowie (1991) on Chariton, and more generally Kuch (2003). van Gennep (1960), at 20. van Gennep’s work has been importantly expanded by Gluckman (1962); Turner (1967) 93–111; Turner (1969). Since the pioneering work of Gluckman and Turner, there has been an enormous quantity of literature on the subject: particularly significant are Vizedom (1976), Droogers (1980) and the various essays in Bianchi (1986). For Greco-Roman applications, Moreau (1992), with extensive bibliography at ii. 297–305; Padilla (1999); Dodd and Faraone (2003). For a critical overview of the applicability of initiation theory to Greek culture, see Versnel (1990) 44–59. See esp. Brelich (1969); Vidal-Naquet (1986a), (1986b), with Ma (1994); Calame (1997). The widespread existence of these rites is argued for by Vidal-Naquet (1986a) 106–28, 129–56 (1986b) and Calame (1997), esp. 89–206. On the other hand, Graf (2003) 20 argues that distinctive ephebic ritual was confined to Crete and Sparta.

First romances

43

focused upon the archaic and classical periods, but much of the evidence is in fact Hellenistic and imperial.83 Rather than seeing the romances as failed Bildungsromane, it is preferable to consider them, with Sophie Lalanne and others, as building upon and developing the classic tripartite passage rite.84 The protagonists are invariably of ‘ephebic’ age, i.e. on the cusp of adulthood.85 Bakhtin’s two ‘biographical moments’ – initial infatuation and ultimate union – represent the pre-liminal and post-liminal rites. What Bakhtin calls the hiatus – that is to say, the substance of the narrative of separation and trial – marks the liminal phase, when the subjects endure marginalisation; the final reunion, ritually marked (in the three later romances) with marriage, coincides with the reintegration of the lovers into their communities as adults. Considering romantic narrative as a variety of the spatio-temporal paradigm of the passage rite not only offers a less anachronistic model than the Bildungsroman, it also allows us to see more sharply what kind of selfhood is being projected, understood in terms not of inner self-realisation (as in the Bildungsroman), but of relationships between individual and community. Two caveats, however, are called for. First, to claim that romances build on passage rite narratives does not mean that they occupy an identical functional role. Sophie Lalanne has argued that the romances, by narrating the socialisation of youths into elite patriarchy within Greek society, simultaneously perform this process, idealising and legitimising the established routes to the cultural and sexual hegemony of elite Greek males.86 This kind of reading, however, pays insufficient attention to the play of narrativity in the ‘liminal’ phase of the novel: in effect, Lalanne presumes (like Bakhtin) that liminality exists solely to be overcome en route to the ending, rather than constituting an experimental space in which the arbitrariness 83 84

85

86

Perrin-Saminadayar (2004) offers a recent survey of findings in relation to the Athenian ephebe lists between the first century bce and the second century ce. Cf. Schmeling (1974) 137–9; Burkert (1987) 66–7; Dowden (1999), (2005); Lalanne (1998) and (2006) (esp. 101–28), a thorough discussion of passage-ritual motifs in the romances; Alvares (2007). For individual studies, see Lalanne (1998) on Chariton; Laplace (1994) on Xenophon (cf. also Bierl (2006), esp. 93), and (1991) on Achilles; Winkler (1990) 101–26, Gual (1992) on Longus; Laplace (1992), Whitmarsh (1999) on Heliodorus. Morgan (1996), 165 n. 7 offers an analysis of the protagonists’ ages: Xenophon’s Habrocomes is sixteen at the beginning (1.2.2), while Anthia is fourteen (1.2.5); Achilles’ Clitophon is nineteen at the outset (1.3.3), Longus’ Daphnis and Chloe fifteen and thirteen at the start respectively (Long. 1.7.1; and two years older at the end). Heliodorus’ Charicleia is seventeen at the end (10.14.4). There is no way of determining the ages of Chaereas and Callirhoe, but the former is designated an ephebe (1.6.5; 8.6.11); Habrocomes and Anthia similarly take part in a procession of ephebes and virgins (1.2–3), as does Heliodorus’ Theagenes (3.3). Lalanne (2006).

44

Part i Returning romance

and plasticity of social roles can be explored (as the cultural anthropologists Victor Turner and Clifford Geertz have argued).87 The wider point, however, is that romances are not simply socially programmatic; they are complex literary narratives, and narratives (as we shall see throughout the course of this book) can model multiple, competing forms of identity. The second point is that the modern concept of ‘passage rites’ is ambiguous, rolling together as it does the ideas of both coming of age and religious initiation. As we shall see in the following chapters, the later romances do make use of religious-initiatory language (er¯os becomes a mystery cult); but there is precious little sign of these motifs in either Chariton or Xenophon, beyond a general resemblance at the thematic level (‘[j]ourneys out and back, descents to suffering and disintegration, ascents to joy and reintegration, these are the stuff of mysteries and of romances too’).88 Even Merkelbach, in the course of his eccentric argument that the romances are religiously functional, exempts Chariton.89 Xenophon, certainly, has a prominent scene of divine intercession, and refers on occasion to the characters’ ‘salvation’ by the gods,90 but there is (pace Merkelbach) no reason to take these features as tokens of a subjacent hermetic religious truth rather than of the general credence in the permeability of the human sphere by divine forces that is almost ubiquitous in the Greco-Roman world. But the use of these religious motifs is also deliberate, reflexive and artful; the romances should be read in terms of ‘ritual poetics’,91 of literary strategy rather than of ‘serious’ religious homiletics.92 The romances are ‘religious’ in the sense that they testify to the presence of numinous forces among us, but ancient polytheism lacked the sharp lines that categorically excluded scepticism, sexuality, mockery and play from religion.93 An excellent exemplification of this comes in the fragmentary Iolaus romance. Here, it seems that Iolaus undergoes initiation into the number of the 87 88 89

90 92 93

See esp. the position statement by Geertz (1979–80), and below p. 214 on Turner. Beck (2003) 150; cited with approval by Zeitlin (2008) 97, who also notes the aptness of Northrop Frye’s description of romance as ‘secular scripture’ ((1976) 97–157). Above, n. 17. Merkelbach’s claim is that historically the romance has its ‘roots’ (Wurzeln) in religious aretalogy ((1962) 333–40, esp. 333), and the texts that we have can be read on two levels, by a general readership and by religious initiates ((1988) 138–9; (2001) 56–59). For critiques of this position see above, pp. 193–204. Below, p. 194. 91 See the essays in Yatromanolakis and Roilos (2004). See Dowden (1996), arguing for Heliodorus’ ‘serious intentions’. Anderson (1982), by contrast, argues for constitutive playfulness. For mocking and obscenity, the classic example is the Athenian gefurism»v (Hsch. G 468, Suda G 212; cf. Plut. Sull. 2.1); for discussion and further literature, see Halliwell (2008) 155–214 on obscene ritual generally, with 169–71 on the gephurismos. For cross-dressing and play as elements of cult, see e.g. Turcan (1992) 226–7.

First romances

45

Galli (castrated servants of Cybele) and dresses as a woman, in order to ‘get a crafty fuck’.94 The joke turns on the bathetic undercutting of the lofty language of initiation with crudely sexual language.95 This is certainly not a ‘realistic’ representation of Cybele’s cult,96 but conversely there is no reason to assume that cross-dressing, role-playing, obscenity, and even sex were thought – whether by initiates or non-initiates – to run against the grain of the cult itself. The Iolaus fragment thus exemplifies particularly luridly the general principle that a Greek romance can be ‘religious’ while also being a sophisticated piece of literary fiction. The crucial point is to distinguish between religious practice, which was in the Greek world typically demarcated from regular life by clear boundaries marking sacred time and space, and religious phenomenology, which permeated almost every area of culture. It makes no sense, then, to see the romances as ‘secularising’97 (or, worse still, as ‘parodying’)98 religiosity, since the intellectual sophistication and play that modern scholars might choose to identify as definitively secular were, in Greek culture, compatible with that religiosity. So, in sum: the claim that passage rites underlie the narrative templates of the early romances commits us neither to the position that the texts therefore mechanistically function to promote established social roles, nor to the position that they therefore exist primarily in order to encode deeper religious truths. symbolic geography 1: xenophon In the early romances of Xenophon and Chariton, ‘abroad’ functions as an absence or negation of ‘home’; and, qualitatively speaking, it represents an inversion (geographic, cultural and ethical) of the patris. Xenophon’s symbolic geography is less nuanced than Chariton’s.99 Ephesus, where the narrative begins and ends, is the centre of the romance, even literally (sited as it is pretty much equidistantly between the easternmost and westernmost 94 95

96 97

98 99

d»lwi . . . bine±n, p. 370.30 SW. Mysteries: [ˆ-]porržtwn (1), mustik»v (14), mustikoÓ (35), t”lei»v . . . g†llov (37); comedy: p”paicen (28), bine±n (30). SW 361–2 place Iolaus in the same ‘criminal-satiric’ bracket as Lollianus’ Phoenician affairs; cf. also 462 on P.Ant. 18 (= SW 464–5). SW 360, who note that elsewhere ‘[i]nitiates are never said to be castrated or to become galli, nor are galli said to be initiated’. E.g. Ker´enyi (1927), at 230; (1971). Similarly, Rohde (1937) and Chalk (1960) argue that Daphnis and Chloe draws freely on elements of mystery religion to create a literary dedication (to Nature, in Rohde; to Eros, in Chalk); Petri (1963) argues much the same of Chariton, though (like his teacher, Merkelbach) he sees the other romances as Mysterientexte. E.g. Gual (1992) 158. H¨agg (1971) 172–5; Lowe (2000) 230–1; Bierl (2006). I begin with Xenophon since his geography is simpler, even though his text may be later (see appendix).

46

Part i Returning romance

extremes of the lovers’ travels, viz. Syria and Italy/Sicily). Conversely, the places visited during the voyage (including Syria, Cilicia, Cappadocia, Phoenicia, Italy, Sicily) signify primarily in terms of otherness, as nothome.100 Xenophontic space, as Anton Bierl felicitously puts it, is the externalised manifestation of the sensation of absence.101 Xenophon offers the best support for Bakhtin’s otherwise overstated claim that all ‘adventures in the Greek romance are governed by an interchangeability of space; what happens in Babylon could just as well happen in Egypt or Byzantium and vice versa’.102 In this space, the characters experience a nostalgia, in the etymological sense: a painful yearning for return. Thus in Tarsus, for example, Anthia meets a shipwrecked Ephesian doctor and ‘enjoyed remembering what it was like at home’.103 For the reader, meanwhile, to remember the Ephesian beginning (as this passage cues us to do) is also to anticipate an Ephesian ending: analeptic memory is merely the converse of proleptic yearning. What are the primary markers of the world abroad? It is not that there is any qualitative cultural difference between Ephesus and other cities (notwithstanding the odd barbarian name like Apsyrtus, or the different language spoken in Cappadocia, 3.1.2). Greek-style civic culture certainly exists in these spaces:104 we read, for example, of an elected official in Cilicia (the ‘superintendent of the peace’),105 and a ‘big, fine city’ in Cappadocia.106 The central difference, in fact, lies not in the cities, but in the places of brooding threat just beyond their walls. Here we find the semi-urbanised villages (4.1.1, 5.2.4–7), and the wilder spaces inhabited by brigands, such as woods (2.11.3–11, 2.13.3) and caves (2.14, 3.3.4, 4.3.6–6.4, 5.2.3).107 It is here that the protagonists are most grievously threatened with murder (2.11.3) and human sacrifice (4.6.4–7). On the seas, outside of these cities (and sometimes within them), gangs of rootless outlaws roam, only occasionally falling foul of the law (2.13.4).108 In narrative terms, these pirates and bandits represent a restless, energetic 100 101 103 105

106 108

See generally Sa¨ıd (1999) 87–8. On the presentation of foreign space as an inversion of Greek see Hartog (1988) on Herodotus, and (2001) more generally; also Malkin (1998) on the Odyssey. Bierl (2006) 75. 102 Bakhtin (1981) 100. ›cairen . . . ˆnamimnhiskom”nh tän o­koi, Xen. Eph. 3.4.3. 104 Sa¨ıd (1999) 95. ¾ tv e«ržnhv . . . proestÛv, 2.13.3; cf. 3.9.5, where the voting is mentioned. Scholars have associated these passages with the office of eirenarch, first attested under Trajan, primarily because this is taken to offer some clue as to the dating of the passage; but we ‘have no right to suppose that our earliest epigraphic testimony is exactly contemporary with the first institution of such an office’ (Bowie (2002) 57). p»lin . . . meg†lhn kaª kalžn, 5.1.1. 107 Sa¨ıd (1999) 87–8. Hopwood (1998) addresses the role of bandits as inversions of normative masculinity. See also below, pp. 217–18.

First romances

47

mobility, the embodied agents of adventure time. Bandits can kidnap others, and move with astonishing speed: at one point, from Tarsus to Laodicea (in Syria), to Phoenicia, then to Egypt and up the Nile almost to Ethiopia, within the space of seventeen lines (4.1). These are the enemies of narrative stasis, ever accelerating and renewing the plot in unpredictable ways. Bandits thus occupy an ambiguous position in the narrative: morally and culturally they invert everything that the protagonists hold dear, but they also embody the narrative energy upon which the romance depends. The major exception to this tendency to code foreign space negatively is Egypt, marked out for special treatment since Herodotus on the grounds of the number of its ‘wonders’ (Hdt. 2.35.1). Here as elsewhere there are certainly bandits and ‘uncivilised’ spaces, such as caves and villages, but Xenophon also shows a keen awareness of the local topography and religion of the Delta region,109 and in particular of Egyptian religion.110 In one extraordinary Egyptian episode, Habrocomes is miraculously saved, firstly, from crucifixion and, secondly, from crucifixion and burning. What makes this passage exceptional is that in general the Greek romancers tend to avoid direct divine intervention, in line with their general commitment to narrative naturalism.111 In this case, in the first instance Habrocomes prays to Helios(-Ra), specifically in his Egyptian guise (‘who dwell in Egypt’, 4.2.4); the god is said to ‘pity’ him (4.2.6), and produces a freak whirlwind that hurls his cross into the Nile. On the second occasion, the Nile waters miraculously rise to put out the flames.112 This exception from the realist rule is perhaps partially legitimised by the allusion to Croesus’ salvation from the pyre (again by Apollo) in Herodotus (1.87.1–2): what is permissible 109

110

111

112

Griffiths (1978) 425–37 and Sartori (1989), contra the more critical claims of Henne (1936) and particularly Schwartz (1985), who claim that the Egyptian landscape is largely traditional. By contrast, Plazenet (1995) 7–9 and Nimis (2004) 46–8 thoughtfully and more plausibly discuss Xenophon’s representation of Egypt as an active engagement with earlier literary and cultural tradition (with earlier literature); see also in this vein Brioso S´anchez (1992). Hence some have claimed that the earliest stratum of Xenophon’s text is an Isiac devotional text: Ker´enyi (1927), e.g. 232–3; Merkelbach (1962) 91–113; Witt (1971) 243–54 adds little; Griffiths (1978). G¨artner (1967) 2074–80, however, effectively rebuts this position, which (for Ker´enyi and Merkelbach at any rate) depends upon the desperate hypothesis of a later, secondary redaction. Morgan (1993) underlines the general emphasis upon reality effects in the romances. The romancers’ practice is also in line with that of New Comedy, where again gods appear in metanarrative roles (notably in prologues), but not as players in the narrative proper: Vogt-Spira (1992) 4–5. See further below, pp. 193–5. A more complex case is Hld. 8.9.10–15, where Charicleia is saved from the blazing pyre: she initially thinks (8.9.16) she owes her salvation to the gods on whom she called (8.9.12), but it is in fact due to the pantarbe stone. Xenophon leaves it unclear to which god Habrocomes prays on the second occasion, but the implication is that it is Helios again (see further below, p. 194). The miraculous nature of the event is indicated by the prefect’s wonder (–qaÅmasen, 4.2.10). On the religious motifs in this entire episode, see Zimmermann (1949/50) 273–7; Merkelbach (1962) 104–6, (2001) 599.

48

Part i Returning romance

to the father of history is implicitly sanctioned for his distant descendants too.113 (It is not impossible, either, that ancient readers would have thought also of Christ’s death on the cross and subsequent resurrection.)114 The crucial point for our purpose, however, is that it is in Egypt that we see the first direct signs of divine intervention in the narrative since the initial references to Eros’ malevolence. Indeed, as a number of scholars have noted, Xenophon’s Egypt is (in line with traditional associations) dense with religious motifs, some explicit (such as the cases above, and the oracle of Apis at 5.4.8–11), some implied (notably Anthia’s burial alive with Egyptian dogs, which, however, refuse to attack her (4.6.3–7) – a scene that evokes both a rebirth ritual and the kind of miraculous salvation that a later Christian martyr might experience in the arena). Particularly remarkable is the role of Isis: she is named as the lovers’ ‘saviour’ (s¯oteir¯ei, 1.6.2) by Colophonian Apollo in the prophecy near the start, and appealed to in this capacity by Anthia when she takes refuge in the goddess’s temple: ‘Queen of Egypt, save (s¯oson) me again; you have helped me so many times before.’115 Unsurprisingly, this Isiac emphasis has been seized upon by scholars wishing to see Anthia and Habrocomes as a devotional work.116 Less credulously, we could take Egyptian Isis as a metastasis of Ephesian Artemis (the two are sometimes assimilated),117 whom the lovers thank with dedications and sacrifices in close association with their ‘salvation’ in the closing lines of the text (s¯ot¯eria, 5.15.2). Even as Egypt is described in terms that mark it traditionally as a site of inversion and alterity, then, the Isiac-salvation theme simultaneously portends the 113

114

115 116 117

Compare Habrocomes’ second prayer (säsai aÉt¼n –k tän kaqestÛtwn kakän, 4.2.8) with Croesus’ (çÅsasqa© min –k tou par”ontov kakoÓ, 1.87.1) – both in indirect speech – and also their outcomes (katasb”nnusi tŸn fl»ga, 4.2.9 ∼ katasbesqnai . . . tŸn puržn, 1.87.2). Habrocomes’ earlier prayer (4.2.5), reported in direct speech, also contains a Herodotean allusion, to the latter’s discussion of the sacred status of those who dive into the Nile (2.90; Zimmerman (1949/50) 275). Xenophon may also be looking sideways towards the Christian martyr tradition, which also makes use of pyre narratives: see ACM 1.15, 2.3, 10.21, 12.4–5. Full discussion of preHerodotean sources for the Croesus story (including Bacch. 3 and Myson’s vase) at Asheri et al. (2007) 141–2. Ramelli (2001) 60: ‘non si pu`o assolutamente affermare che Senofonte avesse in mente la crocifissione di Ges`u’. There are also non-miraculous salvations from crucifixion in Chariton (4.2.7–3.6) and Iamblichus (Photius Bibl. 78a = SW 198–9). Chariton’s crucifixion scene may show some indication of familiarity with the gospels: Ramelli (2001) 36–7). Bowersock (1994) 99–119 also discusses the possible influence of Christ’s resurrection on the romances. å d”spoina A«gÅptou, p†lin säson, ¨i –božqhsav poll†kiv, 5.4.6. On the Isiac details see esp. Griffiths (1978). Ker´enyi (1927) 131–3; Merkelbach (1962) 104–6. For the identification of Isis and Artemis, see Merkelbach (1962) 112–13. The hypothesis of this syncresis (and that of Apollo/Helios/Ra) does make the text more coherent, but cautions against oversystematising Xenophon (e.g. Griffiths (1978) 421–3) are certainly sound.

First romances

49

narrative telos in the civilised Greek space of Ephesus, under the nurturing protection of Artemis. In sum, the religiosity of the text plays a selfreflexively metanarrative role,118 in that the ongoing superintendence of Artemis (∼Isis) and Apollo figures the guarantee of a happy ending to the plot – but without reducing the role of these deities to that of ‘mere’ metaphor for plotting. Another space that occupies a distinctive role in Xenophon’s symbolic geography is Rhodes, where the lovers stop off on both the outward and the return journeys, so that the itinerary becomes chiastic.119 On the outward leg, the Rhodians hold a public festival (1.12.1–2), and Anthia and Habrocomes dedicate an inscription in the temple of Helios (1.12.2). On their return, Anthia dedicates a lock of hair, together with an inscription, in the same temple (5.11.6), again during a festival to Helios (5.11.2); after the recognition, the Rhodians then, in unison as a d¯emos, offer praise to Isis for rescuing and reuniting the pair (5.13.2–3); Anthia and Habrocomes also thank Isis (once again) for their ‘salvation’ (s¯ot¯eria, 5.13.4). The two Rhodian episodes, then, are mirror images of each other, the festivals serving in effect as ritual markers of separation/reincorporation.120 What is more, the second sequence foreshadows the culminating return to Ephesus, where ‘the entire city’ engages in public sacrifice, and a graph¯e (either a picture or a written account) of their story is dedicated in the temple of Artemis (5.15.2). This is conventionally taken as a Beglaubigungsapparat, or device to suggest that the narrative really took place;121 but it also brands the narrative as a whole (or a near-whole)122 as a monument to the salvific power of Artemis (/Isis). The linkage between Ephesus and Rhodes thus figures the dyadic but hierarchical relationship between the

118 119 120

121 122

See esp. Plazenet (1995) 15–16. Bierl (2006) 83, also noting that dreams of female figures occur during the first stay at Rhodes (1.12.4) and just before the second (5.8.5–7). Xenophon also uses the language of ritual in a different metaphorical sense. After returning to Ephesus, his lovers ‘lived the rest of their life together as a festival’ (heort¯e, 5.15.3). This phrase alludes chiastically back to a description of their early married life together, before their parents insisted on sending them away: ‘their whole life was a festival (heort¯e), everything was full of good cheer, and at this point they had forgotten about the prophecies’ (1.10.2). The festival here, however, serves as not so much an instrument of transition as an image of unchanging abundance and happiness: the expression is a popular one (e.g. Plut. De tranqu. an. 477c). The dualit´e s´emantique of the word is discussed by Laplace (1994) 444–5, although her account is distorted by her view that the romance as a whole is an anti-tragic panegyric. See e.g. Morgan (1993) 209; Feeney (1993) 243; Hansen (2003) 308. Sironen (2003) 290–2 offers epigraphic parallels for each of the inscriptions in Xenophon. Hunter (2008a) 268–9 notes that the narrative is not entirely included in the graph¯e, since two sentences’ worth of events remain after the act of dedication.

50

Part i Returning romance

romance’s two principle tutelary deities, (Ephesian/Egyptian) Artemis and (Rhodian/Colophonian) Apollo, the moon and the sun.123 If the impetuous Eros at the start serves partly as a transcription into divine terms of Habrocomes’ own problematic arrogance (see the previous section but one), then the closural Artemis and Apollo serve as more mature, normative models for ephebic behaviour, appropriate to be reintegrated into the community. Shades, then, of the west-Asian/Ptolemaic hieros gamos, the sexual union between brother and sister, that links the cosmic power of the sun–moon pair to its earthly manifestation in a royal couple. In this respect, Anthia and Habrocomes are not merely returning to consolidate their identities: they have become transformed by their journeys, not just in the sense that they have matured as humans, but also in that they have become other, touched and transfigured by the power of cosmic deities. For Xenophon, then, the rituals that conclude the text do not just mark the reintegration of the lovers into their community, nor even proclaim the divine power that oversees the entire process of separation, marginalisation and reincorporation; they also emphasise the godlike nature of the returning couple, by assimilating them to a divine pair. Like the returning victors in Pindaric epinician,124 Xenophon’s returning lovers glow with the magical, transformative power of narrative. symbolic geography 2: chariton Despite the many similarities both general and phraseological between the two texts, the adventure world of Callirhoe is very different. The narrative begins in Sicily; the pirates take her to Ionia, where she is sold to a wealthy Milesian called Dionysius, who becomes her second husband. When Chaereas resurfaces, all the parties travel to Babylon, so that Artaxerxes, the Great King of Persia, can adjudicate the claims. The great king falls for Callirhoe himself; Chaereas is carted off into slavery, before leading a slave revolt in Egypt, and eventually rejoining Callirhoe when he captures a ship with her onboard. Through the snowstorm of different place names (there are, additionally, references to Cilicia, Syria, Armenia, Lycia, Aradus, Cyprus, Crete, Ecbatana, to name but a few), we can identify three principal spatial phases in Callirhoe’s narrative,125 corresponding to her three principal lovers: Sicily (Chaereas), Ionia (Dionysius), Babylon (King 123 124

On Artemis and Apollo as Xenophon’s twin deities, see esp. Zimmermann (1949/50) 267–77. 125 Lowe (2000) 229–30. The classic account is Kurke (1991).

First romances

51

Artaxerxes),126 on a sequential west–east axis. Unlike Xenophon, Chariton (together with his characters) operates with a strong ethical–cultural distinction between Greek and barbarian. The association of Greekness with educated civility (paideia) is shared between Callirhoe (2.5.11; also ‘humanity’, philanthr¯opia) and the narrator (7.6.5); the narrator also refers to the ‘natural servility’ of barbarians,127 and opines sententiously that ‘all barbarians are in awe of the king, and consider him a divine manifestation’.128 Callirhoe thus suffers from progressive deracination on her journey east. A particularly eloquent articulation of this tripartite cultural division comes as she reaches the Euphrates, the boundary of Persia (5.1.3).129 (This geophysical boundary also falls at a textual boundary: the beginning of book 5 marks the opening of the second half of the romance.)130 Full of despair, she reproaches malevolent Fortune, in tragic terms,131 for taking her further still away from home: now you no longer banish me to Ionia. The land you gave me, though foreign, was still Greek, and I had there the consolation of living by the sea. But now you cast me forth from my familiar air, and I am separated from my fatherland by an entire world. You have taken Miletus from me in turn, as before you took Syracuse. Carried off beyond the Euphrates, I, an islander, am imprisoned in the recesses of a barbarian land, where no more sea exists.132 126

127 128

129 130

131 132

Each is pre-eminent in his own context: Chaereas is ‘beyond all’ in terms of beauty (1.1.3); Dionysius is repeatedly marked out as the top Ionian, in station, culture and wealth (2.1.5, 2.4.4, 2.5.4, 2.11.2, 3.6.5, 4.4.3, 8.7.9); Artaxerxes is of course feted for his wealth and power (cf. esp. 6.5.2–5). According to Alvares (2001–2), ‘Chaereas and Callirhoe can be read as a narrative within which three different lovers . . . fail to measure up to what is appropriate to their respective stages, are punished, and, to some extent, are compelled to play their proper erotic roles’ (115). tŸn –mfuton qrhske©an, Char. 7.6.6. katapeplžgasi g‡r p†ntev o¬ b†rbaroi kaª qe¼n faner¼n nom©zousi t¼n basil”a, Char. 6.7.12. See the fuller discussion at Bowie (1991) 188–92, with the additional nuance that the Persians also are said to mistrust the Greeks; also Alvares (2001–2) and Smith (2007) for the themes of Greek liberty vs barbarian oppression. For the Euphrates as a significant boundary between East and West, see also 6.8.6, 7.1.10, 7.2.1, 7.4.11, 7.4.13; and further Lalanne (1998) 546–7. This juncture is marked by a recapitulation of what has been narrated ‘in the earlier part of the story’ (–n täi pr»sqen l»gwi), and a promise that ‘now I shall narrate what happened next’ (t‡ d• —xv nÓn dihgžsomai, 5.1.2). On book divisions and narrative segmentation in Chariton and other Greek romances see Whitmarsh (2009b). Compare e.g. Ajax’ yearning for the ‘holy soil of my native Salamis’ (¬er¼n o«ke©av p”don / Salam±nov, Soph. Aj. 859–60). oÉk”ti g‡r e«v ìIwn©an me fugadeÅeiv. x”nhn m”n, plŸn ëEllhnikŸn –d©douv gn, Âpou meg†lhn e²con paramuq©an, Âti qal†sshi parak†qhmaiá nÓn d• ›xw me toÓ sunžqouv ç©pteiv ˆ”rov kaª tv patr©dov Âlwi dior©zomai k»smwi. M©lhton ˆfe©lw mou p†lin, Þv pr»teron SurakoÅsavá Ëp•r t¼n EÉfr†thn ˆp†gomai kaª barb†roiv –gkle©omai muco±v ¡ nhsiätiv, Âpou mhk”ti q†lassa, Char. 5.1.5–6. This passage is perceptively discussed by Daude (1990) 86–8, who notes the resemblance to the Acheron (only to be crossed once), the symbolic status of the Euphrates

52

Part i Returning romance

Callirhoe’s tripartition of the world is highly visible here, with Ionia as the hybrid centre, ‘foreign, but Greek’: a paradox she expresses through a polarising jingle, xen¯en men, pl¯en Hell¯enik¯en. She also expresses the division in terms of her affection for the Mediterranean/Aegean, the ‘Greek sea’ as she has earlier put it,133 reversing the usual model of sea as terrifying and land as comforting: Sicily is preferred as an island. Ionia is next best, as a coastal mainland. Persia is worst of all, (imagined as) landlocked, an English metaphor that captures her own phrasing: ‘imprisoned in the recesses’ images Persia as a jail where she will be constrained, or perhaps more mythopoetically the Cyclops’ cave,134 or even the infernal regions.135 Even though there is no explicit allusion here, Callirhoe surely has in mind the famous cry of relief ‘the sea, the sea’, uttered by (Athenian) Xenophon’s Greek mercenaries who have felt themselves trapped in Asia.136 The metaphors she uses of this inland imprisonment suggest suffocation, perhaps even (an aggressive paradox) drowning: she imagines herself deprived of ‘air’ down in the ‘depths’ of the Asian continent.137 Ionia, as we have noted, is presented as a liminal space, a meeting-point between Orient and Occident. On the one hand, we do find the traditional references to Ionian luxury and the corruptive influence of the East (1.11.7, 5.10.7–8); but this is also where Callirhoe meets Dionysius, nonpareil in the cardinally Greek values of civilised education (paideia: 1.12.6, 2.1.5, 2.4.1, 2.5.11, 3.2.6, 4.7.6, 5.5.1, 5.9.8, 8.5.10) and humanity (philanthr¯opia: 2.2.1, 2.5.3–4, 2.5.11, 2.7.2; in another life, in another romance, he might have been quite a catch). Ionia is Janus-faced, looking both inland to the Persian East (as the pirate Theron comments, ‘royal riches flow in from all over Asia’, 1.11.7) and out to the Greek-dominated cultural world of the Mediterranean. This is what makes it a site of cultural fluidity and dynamic action, and hence so central to romantic action. For the romancer, the liminality of Ionia, equipoised between East and West, makes it a place of opportunity, brimming with narrativity: it is here that the complex moral questions begin to pose themselves, for Callirhoe and Dionysius alike. That Chariton has chosen this role for Ionia, of all places, is surely no coincidence. Miletus, Dionysius’ homeland and the location of most of the Ionian narrative, is the city closest to Aphrodisias, Chariton’s own 133

134 136

as the boundary of the Roman empire during Chariton’s time, and its status within Chariton’s romance as an axiological boundary between barbarian and Greek, slave and free. qal†sshv ëEllhnikv, 4.7.8 (the narrator speaks here, but apparently focalising Callirhoe’s thoughts). Bowie (1991) 189 n. 17 notes the use of this phrase at Hdt. 5.54.2, Thuc. 1.4.1, and Arr. Anab. 2.25.1, 5.1.5. 135 Cf. Hes. Th. 119 (mucäi gv). Which also has a muc»v: Hom. Od. 9.236. 137 Seas can be imaged in terms of mÅcoi: see e.g. [Aesch.] PV 839. Xen. Anab. 4.7.24.

First romances

53

(which, of course, did not exist at the time when the narrative is set, at least not as a Greek polis): the former lies on the edge of Caria, at the mouth of the Great Maeander, the vital inland trade route in the basin of which Aphrodisias lies.138 Chariton’s Miletus, with its active cult of Aphrodite, seems to serve as an ancient metastasis of the modern city of Aphrodisias.139 The crucial symbolic role of Ionia in the romance, then, as a place of transition and confluence, figures Chariton’s own identification of this romance, and perhaps the romance in general, as an Ionian form. Callirhoe’s symbolic geography is also (as a number of scholars have argued) an allegorical response to the experience of empire, mapping out a secondary story of both benign and malign political hegemony in the background of the primary, erotic narrative.140 This is underlined by the narrative setting, which focuses upon the victorious Syracusans in Sicily in the aftermath of the attempted Athenian invasion of 416 (the tragic centrepiece of Thucydides’ narrative of Athenian decline into deluded arrogance). From the very start – Callirhoe is programmatically identified as the daughter of ‘the man who defeated the Athenians’ (1.1.1)141 – there is an air of overweening menace from alien powers abroad. Even the nonSyracusan (?) brigand Theron characterises Athens in negative terms, for the ‘inquisitiveness’ (polupragmosun¯e) of the people, who are ‘gossipy and litigious’ (lalos kai philodikos); ‘sycophants’ abound at the harbour, and ‘vile suspicion will overtake those wretches’ if his crew land there.142 Chariton’s presentation of Athens as a place of malevolence and iniquity, rather than the hub of Greek civility, is all the more striking for its variance from the usual picture painted by early imperial writers.143 Like alienation, imperialism intensifies along the West–East axis: Persia is to Athens what Athens is to Syracuse. Callirhoe tells the eunuch Artaxates that the latter is a ‘city that not even the Athenians conquered – the Athenians who conquered your “great king” at Marathon and Salamis’.144 138

139 140 141 142

143

144

On the importance of the Maeander valley to Aphrodisias, see Reynolds (1982) 3, 31, 81. In late Roman times, it seems, seats were reserved in the theatre of Aphrodisias for Milesian spectators (Rouech´e (1993) 90). Ruiz Montero (1989) 126; Jones (1992) 162–3; Alvares (2001–2) 126–7. Alvares (2001–2); Connors (2002), (2008); Schwartz (2003); Smith (2007). Other allusions to the campaign at 6.7.10, 8.6.10. Ëpoy©a katalžyetai ponhr‡ toÆv kakožqeiv, Char. 1.11.6. Theron also refers to the threat posed by the Areopagus court, and its ‘archons more severe than tyrants’ (Šrcontev tur†nnwn barÅteroi, Char. 1.11.7). On Chariton’s representation of Athens, see Oudot (1992) 101–3; Alvares (2001–2) 119–20; Smith (2007), esp. 50–98; for the generally positive view of Athens in the imperial period, see Bowie (1970) 195–7; Smith (2007) 23–49. p»lewv . . . ¥n oÉk –n©khsan oÉd• %qhna±oi o¬ –n Maraqäni kaª Salam±ni nikžsantev t¼n m”gan sou basil”a, Char. 6.7.10.

54

Part i Returning romance

Syracuse, Athens and Persia form a rising tricolon of (ineffective) militarism.145 The might and extraordinary efficiency of Persia’s army is expressed in the narrator’s description of the aftermath of the Egyptian rebellion. ‘Everyone agreed on speed, and not to delay even for one day if necessary’,146 but in any case ‘the mobilisation of forces is very swift for the Persians’.147 The following description catches itself savouring the awesome spectacle of empire in action, and invites the reader to do the same. ‘It had been decreed since the time of Cyrus’, we are told, ‘which tribes should provide cavalry and how much, which infantry and how much, how many archers each should provide and how many regular and scythe-bearing chariots, where the elephants should come from and in what numbers, from whom the money should come, in what form and what quantities’.148 The narrator’s emphasis upon precise naming, calibration and quantification temporarily aligns literary with military strategy (as in the passage’s distant ancestor, the Iliadic catalogue of ships), performing on the king’s behalf the rhetoric of empire. This is romance sloping into epic, which ‘loves a parade’.149 The tricolon, however, rises in both directions simultaneously: the states become mightier towards the East, but more heroically resistant towards the West. The Sicilians Chaereas and Polycharmus join the rebellion against Persia and its ‘tyrant’ to demonstrate ‘that two wronged Greeks aggrieved the Great King in return, and died like men’.150 The rebellion becomes a testing-ground for manly virtue.151 Chaereas ‘duplicates the naval victory at Salamis, Xenophon’s retreat with the Ten Thousand, and Alexander the Great’s conquest of Tyre’.152 As with the Syracusan defeat of Athens, as with the Athenian defeat of Persia, so the rebellion against the imperial order is successful: military history in Chariton is distilled into a series of dramatisations of the simple ethical truth that the noble and free fight better. Conversely, Chariton’s portrait of Persia, as scholars have been quick to recognise, is traditional in its orientalising portrayal of a luxurious but 145 146 147 148

149 150 151 152

For similar tricola, see 2.6.3, 5.8.8, 7.5.7–8; Smith (2007) 92–3. psi . . . ¢reske t¼ speÅdein kaª mhd• m©an ¡m”ran, e« dunat»n, ˆnabal”sqai, Char. 6.8.5. ç†isth d ì –sti P”rsaiv ¡ paraskeuŸ tv dun†mewv, Char. 6.8.7. sunt”taktai . . . ˆp¼ KÅrou . . . po±a m•n tän –qnän e«v p»lemon ¬ppe©an kaª p»shn, t©nav d• tox»tav kaª p»sa —k†stouv Šrmata yil† te kaª drepanhf»ra, kaª —l”fantav ¾p»qen kaª p»souv, kaª cržmata par ì æntinwn, po±a kaª p»sa, Char. 6.8.7. Quint (1993) 31. dÅo í Ellhnev ˆdikhq”ntev ˆntelÅphsan t¼n m”gan basil”a kaª ˆp”qanon Þv Šndrev, Char. 7.1.8. The rebellion has been connected with both the revolt against Persia in 360 bce (Salmon (1961)) and events in the Roman era (Alvares (2001)). Alvares (2007) 13; cf. Lalanne (2006) 91–2, noting the language of andreia in books 7–8; also 156–9.

First romances

55

despotic state (even though he may draw on an unexpectedly wide range of sources).153 As one scholar puts it, ‘Chariton deploys all the standard elements of the Persian mirage: luxury, prostration, harem life, eunuchs, satraps, court intrigue, hunts, magi, the paradeisos’.154 It is in particular the steepling power, with its vertiginous allure, that overhangs Chariton’s Persia. There is in Persia none of the honourable competition between equals that characterises Greek states like Syracuse, only the master–slave paradigm. At the apex sits the king, ‘to whom’ (opines his favoured eunuch) ‘all fine things are enslaved: gold, silver, clothing, horses, cities, nations’.155 This pattern is reproduced further down the pyramid, so that (for example) the satrap Mithridates is the slave of the king (5.2.2), and Dionysius, the pre-eminent citizen of Miletus, is the slave of Mithridates (6.7.9). This model of human relationships, of course, demands to be set against the norms of Greek society, with its supposed freedom and civic unity. The othering of Persia is compounded by numerous echoes of, allusions to and quotations from the Iliad, in particular the explicit comparisons of Callirhoe with Helen, and the implicit associations of the Egyptian revolt with the Greek mission: all of these serve to assimilate Persia to Troy, the enemy that must be defeated.156 The stereotyping of Persia, as the model of a state in which interpersonal relationships are warped by vast differentials of power, gains new urgency when considered in relation to the Roman imperial context for which Chariton was writing. As we have seen, the author’s home town of Aphrodisias benefited immensely during the early first century from the patronage of the Julio-Claudian emperors, who in commemorating its patron goddess were also honouring their own ancestor. Aphrodite is also the presiding deity of Callirhoe, and given a decisive role in the management of the harmonious ending.157 At one level, then, Chariton’s celebration of Aphrodite as a benevolent power reuniting the lovers despite their trials and suffering offers itself as an analogy at the divine level to the Roman emperor’s governing of the empire at the political. The imperial resonances are strengthened thanks to the choice of Syracuse as the lovers’ homeland. Centring the world in Sicily involves decentring 153 155 156

157

154 Schwartz (2003) 378. Bowie (1991) 188–95; Baslez (1992) emphasises the range of sources. æi t‡ kal‡ p†nta douleÅei, crus»v, Šrgurov, –sqžv, ¯ppoi, p»leiv, ›qnh, Char. 6.3.4. See, however, below on the ambiguous presentation of the eunuch Artaxates. Callirhoe as Helen: 2.6.1, 5.2.8, 5.5.9. Persian sexual ethics are assimilated to Trojan when the narrator describes these barbarian nature as ‘woman-mad’ (gunaikoman”v, 5.2.6), invoking the phrase Hector uses of Paris (gunaiman”v, Hom. Il. 3.39). On the Iliadic resonances of the Egyptian uprising see, De Temmerman (2009a) 255–7. Above, n. 12.

56

Part i Returning romance

the Greek mainland, shifting it further towards the barbarian East – a symbolic construction of the world that arguably reflects the Italocentric imperial mapping of Rome more than traditional Greek ideas. Moreover, as Catherine Connors has noted, Rome’s relationships with the Greek world had long been played out through its treatment of Syracuse (particularly since the sack by Marcellus in 211 bce); in Chariton’s own day, there was reason to see it as flourishing thanks to Julio-Claudian repression of piracy.158 If these features emblematise the positive aspects of empire, then conversely Achaemenid Persia represents the inverse: and, via an obvious d´ecalage, also the Parthian kingdom that was seen as such a threat to Rome in Chariton’s own day (when his hometown of Aphrodisias was still remembered for its resistance to Parthian overtures during the Triumviral period).159 It is, however, impossible to insulate Persia from Roman associations. Greeks of the imperial era like to create fleeting, suggestive parallels between Persia and Rome: hence the occasional use of the Persian word satrap to denote Roman provincial governors,160 and Plutarch’s famous suggestion to Menemachus of Sardis that orators should avoid inflammatory references to ‘Marathon, Eurymedon and Plataea’ (Political advice 814c), which suggests that analogies could be and were drawn in rhetorical contexts.161 As one scholar has observed, Chariton’s court scene, with its combination of petition and documents, is remarkably evocative of practice at the imperial court at Rome.162 In fact, despite the othering described above, there is plenty in Persia that seems uncannily resonant to a reader familiar with Roman imperialism.163 At the level of empire, Persia is ruled politically, militarily and juridically by one man, who mandates the rule of provinces to individual governors – just like Rome. At the religious level too, no attempt has been made to differentiate Persian religious practices from Greco-Roman: the pantheon is identical, as is the use of festival (with the Greek features of garlanding, incense and wind instruments) and sacrifice as the primary means of mediation between the two.164 Even more disturbing to any comfortably compartmentalising reading in terms of alterity is the often positive representation of the Great King. 158 160 161 162 163 164

159 On Parthian features in Callirhoe’s Persia, see Baslez (1992) 203–4. Connors (2002). Dio Chr. 7.66, 7.93, 47.9, 49.6, 50.6; Philostr. VS 524. See further Jones (1971) 113–14 and Whitmarsh (2005a) 66–7. Schwartz (2003) 382–5; also Alvares (2001–2) 120–3. Cf. esp. Alvares (2001–2) 120–3, noting particularly Mithridates’ Roman-style plantation (4.2.2), the crucifixion of slaves and the presence of freedmen (5.4.6). Pantheon: Helios (6.1.10), Eros (6.2.4, 6.3.2), Aphrodite (6.2.4), Zeus (6.3.2). Festival: 6.2.3–4. Sacrifice: 6.2.2, 6.2.4.

First romances

57

Like Dionysius, he is a strikingly sympathetic figure, whose struggles with his passion for Callirhoe are offset by a strong awareness of his social and ethical obligations. Particularly telling is his resistance to the allures of power. There are, certainly, suggestions of divinisation: the people ‘consider the king a god manifest’,165 and his eunuch Artaxates presents him, albeit problematically, as superhuman: ‘you alone, master, can overpower even a god’ (i.e. Eros).166 The king himself, however, resists this: despite some hints (he claims Helios as his ancestor (6.1.10), and refers mysteriously to ‘the royal gods’ (6.2.2)), he acknowledges the superior power of Eros (6.3.2), elsewhere enjoining a submissive ‘piety’ before deities in general (eusebeia, 6.2.2). Given that living Roman emperors were worshipped as gods in the eastern empire,167 and that this was a phenomenon that some educated Greeks of the early imperial period found awkward to reconcile with the traditional resistance to the superelevation of mortals,168 the representation of Artaxerxes as moderate in the face of temptation looks like an intervention on Chariton’s part in a contemporary debate on the divinity of the emperor.169 More generally, Chariton’s king insists on abiding by the law and the ‘justice that I practise in all things’170 – again, in the context of his eunuch’s encouragement to break the law (by raping Callirhoe). Here too, he seems to embody one aspect at least of Greco-Roman kingship ideals, here the principle that the king should be ‘law incarnate’ (nomos empsykhos).171 These debates between king and eunuch serve as a softened, eroticised version of the ‘constitutional debate’ of Herodotus 3.80–2, conducted (appropriately) by three Persians.172 Unlike Herodotus, however, Chariton here represents kingship as the only political option; the only question is good or bad. The eunuch’s vision of kingship as limitless power represents the alternative road that Chariton has not taken, the extreme othering of Persian kingship that the text represses. The near-homonym Artaxates can be seen as Artaxerxes’ shadowy double, onto whom all monarchy’s negative potential for abuse and arrogance has been shunted. In the final analysis, the king embodies 165 166 167 168 169

170 172

qe¼n faner¼n nom©zousi t¼n basil”a, Char. 6.7.12. dÅnasai . . . æ d”spota, sÆ m»nov krate±n kaª qeoÓ, Char. 6.3.8. Price (1984) is the classic account, emphasising continuity with Hellenistic practice. See Bowersock (1973) on Greek intellectuals’ awkwardness in relation to imperial cult. Comparable is Plutarch’s (extraordinary!) claim that Alexander adopted divinisation ‘moderately and sparingly’ (metr©wv kaª Ëpofeidom”nwv) before the Greeks (Alex. 28.1): Whitmarsh (2002a) 191. 171 Gigante (1993). dikaiosÅnhv ¥n –n Œpasin ˆskä, Char. 6.3.8. Also imitated at Philostr. VA 5.27–38: see Whitmarsh (2001a) 230–8, with further literature on both Philostratus and Herodotus (and Pelling (2002)).

58

Part i Returning romance

not so much the pathological despotism of the East as the kind of benign imperialism that Greeks can do business with.173 There is thus a paradox underlying Chariton’s geography. At one level, the text rests upon and consolidates a simple opposition between self and other. This opposition offers itself to quick decipherment, as a political allegory opposing Rome’s benevolent rule (symbolised by Aphrodite) and the threat of Parthia, corrupt and oppressive. This simple reading, however, is offset by another, whereby the further we get from Italy, the closer we get to Rome, or rather ‘Rome’ – the idea of absolute power concentrated in one man, of absolute dominion over imperial provinces, and of benign imperialism. Callirhoe models the neat Hellenocentrism that underlies Anthia and Habrocomes, as warped by Aphrodisian complicity with Roman providentialism. From those like Chariton’s compatriots who attempt to shape their own cultural patterns to its ideology, empire demands an impossible feat of intellectual contortion: an accommodation must be found between the face-to-face community’s traditional sense of its moral superiority over imperialist threat (epitomised not only by Persia but also by Athens) and a recognition that, by Chariton’s time, benign despotism is the best one can hope for. healing the rift The end of the romance marks the reintegration of the lovers into the home community. Xenophon, as we have seen above, articulates this transition ritually: both on Rhodes and in Ephesus, the lovers are welcomed back with pandemic festivals, in thanks for their salvation. In Chariton, typically, the articulation is by contrast political. The lovers are acclaimed by the Syracusan populace both on the shore (8.6.7–8) and in the theatre (8.7.1–2); the civic spaces are repeatedly claimed to be ‘full’, the superabundance marking the people’s zeal and concord.174 The people’s investment 173

174

‘In defeating Artaxerxes’ forces but returning his Queen, Chaireas, like Hermocrates before him, has reached a modus vivendi with a dominant power, as Callirhoe also does in her own way’ (Alvares (2007) 14). In particular, the Persians and the Sicilians are united by their common enemy Athens: 2.6.3, 5.8.8, and Smith (2007) 98. ‘The entire harbour was filled with people’ (pv ¾ limŸn ˆnqrÛpwn –neplžsqh, 8.6.5); ‘The harbour was full ’ (¾ limŸn –plhroÓto, 8.6.10); ‘The people shouted in unison’ (ˆqr»on . . . t¼ plqov ˆneb»hsen, 8.7.1); ‘The theatre was full of the talk of women and men’ (l»gou . . . –plhrÛqh t¼ q”atron ˆndrän te kaª gunaikän, 8.7.1); ‘Prayers from everyone followed’ (eÉcaª par‡ p†ntwn . . . –phkoloÅqhsan, 8.8.12); ‘The people approved . . . the people shouted . . . ’ (–peufžmhsan ¾ dmov . . . ¾ dmov –peb»hsen, 8.8.13). Incidentally, Chariton’s theatre is conceived of as a political rather than a religious space: cf. 8.7.1 (–kklhs©an), 8.8.14 (yžfisma –gr†fh).

First romances

59

in this narrative is also underlined by the claim that they were happier than after the defeat of the Athenians (8.7.2; cf. 8.6.2, 8.6.10, 8.6.12). The reunion of Chaereas and Callirhoe reunifies the city, an analogy that Chariton underlines by explicitly gendering the populace: ‘then they [the people] were sometimes split, the men exalting Chaereas, the women Callirhoe – and sometimes then they jointly (koin¯e) exalted them both again, which was more pleasurable’.175 The harmonious alignment of women and men (distantly invoking the joyous gender reintegration at the conclusion of Aristophanes’ Lysistrata) points up the positive contrast with Babylon, where (we are told) before the trial ‘the barbarian populace was split’ along lines not only of gender but also of class.176 The reference to ‘pleasure’ also directly links the union of the genders at the civic level to the personal relationship between Chaereas and Callirhoe, recalling the authorial recapitulation at the start of this book (8), promising that this one will be ‘the most pleasurable to readers’ in that it offers ‘legitimate love and marriage in the eyes of the law’.177 The reunion of Chaereas and Callirhoe figures the unity of the d¯emos. Book 8 of Chariton, indeed, is strongly closural.178 It begins, as we have noted, with a major recapitulation, the second of the romance.179 The first comes at the beginning of book 5, the half-way point in terms of the book structure, and also the point in the narrative where Callirhoe crosses the boundary between East and West, the Euphrates. The recapitulation serves as a narrative pivot, waymarking the text: it both summarises events ‘that have been shown in the story so far’180 and looks forward, promising that ‘I shall now narrate what happened next.’181 This second recapitulation also serves to segment the text, by explicitly hiving off the eighth book as closural: it distinguishes between the ‘gloomy events of the foregoing story’182 and what will happen in ‘this, the final book’,183 which he promises 175 176 177 178 179 180

181 182 183

e²ta pot• m•n –sc©zonto, kaª o¬ m•n Šndrev –pžinoun Cair”an, a¬ d• guna±kev Kallir»hn, pot• d ì aÔ p†lin ˆmfot”rouv koiniá kaª toÓto –ke©noiv ¤dion §n, Char. 8.7.2. –sc©sqh . . . t¼ plqov tän barb†rwn, Char. 5.4.1–4, at 5.4.1. to±v ˆnaginÛskousin ¤diston / ›rwtev d©kaioi . . . n»mimoi g†moi, Char. 8.1.4. Fusillo (1997) 215–16; see further below, pp. 182–4. For fuller discussion of the recapitulations and their relation to the book divisions see Whitmarsh (2009b), with literature. –n täi pr»sqen l»gwi dedhlÛtai, Char. 5.1.2. The phrase is borrowed from the tags that Xenophon of Athens (or an editor) positions at the start of books 2–5 and 7 of the Anabasis (Perry (1967) 358 n. 16). In Xenophon, ¾ pr»sqen l»gov means ‘the previous book’, while in Chariton it means ‘the story so far’. t‡ d• —xv nÓn dihgžsomai, 5.1.2. tän –n to±v prÛtoiv skuqrpän, 8.1.4; cf. –n täi pr»sqen l»gwi, 8.1.1, skuqrwp»n, 8.1.2. t¼ teleuta±on toÓto sÅggramma, 8.1.4.

60

Part i Returning romance

to be the most pleasurable (h¯ediston) for his readers.184 Unlike the parallel passage in book 5, where readers are left to guess ‘what happened next’, Chariton here leaves his reader in no doubt about the events to follow, explicitly pre-empting the ‘pleasurable’ ending: ‘legitimate passions and marriage in the eyes of the law’.185 This disclosure is itself a closural device, signalling that suspense no longer has a role to play. The metanarrative description of the movement from the liminal to the closural phase is reinforced within the narrative at three levels. Cosmically, the president deity (as we shall see in greater detail later in this book) changes from malevolent Fortune to benign Aphrodite. Socially, ‘reconciliation’ betokens (again as in Aristophanes’ Lysistrata, where Diallag¯e is personified) the beginning of the closing phase of the narrative, as rifts are mended and anger is healed – under the tutelage of the patron goddess of Chariton’s home-city of Aphrodisias, and the ancestor of the imperial dynasty of the day.186 Psychically, finally, the passing of Aphrodite’s anger betokens the maturation of the lovers, as they set aside socially destructive emotions (particularly Chaereas, whose ‘jealousy’187 is recalled here). Indeed, all three levels are interrelated: the lovers, the city and Aphrodite, all master their passions. This technique, whereby gods serve as external manifestations of internal passion, we have already seen in Anthia and Habrocomes: the changing identity of the key gods (haughty Eros > serene Apollo/Artemis) also indicates growing up. Again as in Xenophon, anger is the dominant divine motif: Xenophon’s Eros ‘rages’ and ‘grows angry’,188 and that the narrative ahead is conceived of by Eros as a ‘great penalty’ for Habrocomes’ arrogance.189 This discourse of divine anger (transferred, as we have already observed, from epic) serves to construct the forthcoming narrative as a deviation from anticipated equilibrium at the levels of divine and social order. It also looks towards a happier end: anger cannot sustain itself indefinitely in narrative, it exists to be spent. Divine wrath, thus, portends a circumscribed period of punishment, but thereafter a restitution of the normal psychological and social order. closing the case This sense of restitution of order is an integral part of the return narrative. The Greek romance is fundamentally preoccupied with closure: 184 185 187 188 189

On the segmentary role of the recapitulations, see also Reitzenstein (1906) 95–6 and esp. H¨agg (1971) 246–52, noting phraseological parallels between the two passages (246–7). Text at n. 177. 186 On Charitonian anger management, see Scourfield (2003). Reemphasised in a passage focalised by Callirhoe, at 8.4.1. mhnii, Xen. Eph. 1.2.1; Ýrg©zeto, 1.4.5. meg†lhn tv Ëperoy©av . . . timwr©an, Xen. Eph. 1.4.5.

61

First romances

narratologically, aesthetically, psychologically, ideologically. We shall return to this point more fully in Chapters 5 and 6, but for now it will be important to make two general points about the first-century romances. The first is that closure can never, no matter how hard one tries, be total. Closure it not simply the resolution of issues, but the imposition of resolution; and when, as readers, we perceive this imposition to be forcible or contrived we protest all the more (as, for example, Aristotle did with the Euripidean deus ex machina).190 What is more, the pleasure of the ‘happy ending’ – the h¯edon¯e predicted by Chariton’s narrator – is for the reader deeply ambivalent, since we will be aware how much guilty pleasure we have found in the yearnings and moral cruces enacted in the main narrative. In D.A. Miller’s ascetic account, the very precondition for narrative – ‘the narratable’ – lies in the uncertainties and emotional cruces that are precisely erased by closure. In other words, closure is not part of the narrative, only a device to indicate its absence: ‘The closural settlement accommodates the narratable only by changing its status, that is, by putting it in a past perfect tense and declaring it “over”. Closure can never include, then, the narratable in its essential dimension: all suspense and indecision.’191 Closure, on this reading, does not resolve the meaning of a narrative, because the meaning of narrative lies in its very indeterminacy; what closure actually does is cancel narrative by transubstantiating it. Even if this is a wilfully restricted definition of what narrative ‘is’, it nonetheless effectively captures the paradox of closure, which stands both within and without the narrative, simultaneously completing and liquidating it. These issues matter, because (as, again, we shall see in greater detail in Chapters 5 and 6) they determine whether we see the romances as fundamentally conservative expressions of civic ideology or as ludic experiments with narrative and social possibilities. The best answer to this question comes from Stanley Fish, who emphasises that the meaning of a literary work is not realised in a single cognitive act, but gradually composed over time: Everything depends on the temporal dimension . . . In a sequence where a reader first structures the field he [sic] inhabits and then is asked to restructure it . . . there is no question of priority among his structurings; no one of them, even if it is the last, has privilege; each is equally legitimate, each properly the object of analysis, because each is equally an event in his experience.192

190

Arist. Poet. 1453a37–1454b6; Lowe (2000) 58.

191

Miller (1981) 98.

192

Fish (1976) 474.

62

Part i Returning romance

If we see meaning-making in this way as a process, then closure, for all that it does suggest a final resting point, does not erase the memory of all the steps that led there: in the end we settle on the more optimistic reading – it feels better – but even so the other has been a part of our experience, it means. What it means is that while we may be able to extract from the poem [or narrative] a statement affirming God’s justice, we are not allowed to forget the evidence (of things seen) that makes the extraction so difficult.193

Closure, then, is not a state but a process. This is crucial. In Xenophon, and particularly in Chariton, considerable attention is given to the ways in which life as experience (open, unresolved) is translated into life as narrative (closed, shaped, heavy with meaning): to the transition from Erlebnis to Erz¨ahlung (to use the terms popularised by Walter Benjamin). The first ‘readers’ of the narrative are the characters who live through it, and it is their attempts to give shape to their experiences to which we now turn. In Xenophon, the act of narrative control comes in the form of a ‘graph¯e of all that they had endured and all that they had done’, deposited in the temple of Artemis.194 Graph¯e can mean either a ‘written record’ or a ‘painted depiction’, but the sentence obviously advertises its availability to a self-reflexive interpretation, whereby the dedication actually constitutes the book we are reading now.195 Xenophon’s is a ‘self-begetting’ text.196 Anton Bierl notes additionally that Anthia and Habrocomes effectively translate their story generically too, so that it becomes an aretalogy: events that were experienced phenomenologically as painful are now metamorphosed into a testimony to the goddess’s benevolence.197 Jason K¨onig notes one further translation, from oral into written narrative. The story is now congealed and consigned to the past, and hedged with images of death: ‘The closing paragraphs of the work offer a fantasy of fulfilment, but they also have overtones of finality, for example in the atmosphere of death in the final lines, where we hear that the main characters lived out “the rest 193 194

195

196

Fish (1976) 470. tŸn grafŸn . . . p†ntwn Âsa te ›paqon kaª Âsa ›drasan, Xen. Eph. 5.15.2. Ephesian Artemis’ temple seems to have been the final resting place for numerous artistic creations, including Heraclitus’ On nature (Diog. Laert. 9.6) and the ‘psaltery’ of Alexander of Cythera (FGrH 275 F 83 = Ath. Deipn. 183c); see also the following note, on Apollonius, King of Tyre. Hansen (2003) 309 notes the ‘light pseudo-documentarism’. See esp. Hunter (2008a) 267–9, emphasising the parallels not only with Longus, but also with Apuleius’ Metamorphoses (where the afterlife of the narrative as a text is frequently envisaged); and with Apollonius, King of Tyre, the B and C redactions of which have copies of the story deposited in the king’s own library and in the temple of Diana at Ephesus. Kellman (1980). 197 Bierl (2006) 80–1.

First romances

63

of their lives” in Ephesos, with mention also of the building of tombs and memorials.’198 In metamorphosing narrative into something bounded, with a fixed meaning – the city’s patron goddess will save us in times of adversity – Anthia and Habrocomes also neutralise the energy that sustained that narrative, and in a sense destroy their own vitality: all that remains now is to die. Anthia and Habrocomes is ‘finally’ about the salvific power of civic religion. In Chariton, characteristically, the emphasis is partly upon community, but partly also on stable psychology and harmonised social relations. When Chaereas and Callirhoe recognise each other, the two lovers recount their stories to each other (8.1.14–17). This reinforces the sense of movement towards narrative resolution, partly in that it echoes the narrator’s explicitly closural recapitulation at the start of the book (discussed above), partly in that it reworks the reunion of Odysseus and Penelope (Od. 23.310– 41 – just after the point (296) where Aristophanes and Aristarchus located the end of the text), but mostly in that it marks the conciliation of the two lovers. Whereas the narrator’s recapitulation has ‘no organic function in the action itself’,199 Callirhoe’s story of her remarriage to Dionysius is a tense moment, threatening as it does to destabilise her relationship with Chaereas: when she comes to Miletus in the story, ‘she fell silent, in embarrassment’.200 Chaereas, however, is all contrition; he exhorts her to tell the story and apologises for his earlier anger. The tension inherent in this act of recapitulation is thus defused (even though both Callirhoe and the narrator recall his ‘innate jealousy’ (emphutou z¯elotupias)).201 Callirhoe finds a speaking cure to her trauma: she overpowers her aid¯os through narrative recapitulation, by becoming a narrator of her story in her own right. The reliving of narrative, as Freudian critics emphasise, can engender a sense of mastery over traumatic events.202 In a parallel scene, Chaereas recounts his story to the Syracusan public in the theatre (8.7.9–8.11) – again, a performance that retraces, and in a sense (to be discussed below) reconstitutes, the narrative that it caps. The different audiences are significant: whereas Callirhoe’s primary human relationship is with her husband, the now-mature Chaereas needs to claim 198 200

201 202

K¨onig (2007) 18. 199 H¨agg (1971) 251. –siÛphsen a«doum”nh, Char. 8.1.15, echoing her reluctance to tell her story to Dionysius at 2.5.6–7. In that context, a brief recapitulation is wrung out of her against her will, testimony not to her mastery of the situation but to her powerlessness (2.5.10–11). Callirhoe’s narrative to Chaereas is also parallelled by Stateira’s to Artaxerxes (8.5.7–8). Callirhoe presently writes to Dionysius without her husband’s knowledge, ‘knowing his innate jealousy’ (e«du±a . . . aÉtoÓ tŸn ›mfuton zhlotup©an, Char. 8.4.4). Brooks (1984) 99–100, influentially reworked for the Aeneid by Quint (1993) 50–96.

64

Part i Returning romance

his identity as a citizen before the populace. (Callirhoe, in fact, has been led away from the theatre: 8.7.3). Chaereas’ account, which lays particular emphasis upon his martial exploits, dramatises his newfound manhood on the public stage (literally so, in the city’s theatre).203 If this self-aggrandising tale of exploits casts him as the Odysseus of the Phaeacian apologoi, then the first appearance before the city also evokes Telemachus’ maiden speech in the Ithacan assembly, an event that marks the latter’s emergence into manhood.204 The coming-of-age theme is, indeed, emphatic. Chaereas’ initial reluctance to relate his whole story is presented as resulting from the shyness of youth: ‘Chaereas hesitated, embarrassed (aidoumenos) as you would expect by the many events that had occurred against his will.’205 Hermocrates’ subsequent advice to him – ‘do not be at all embarassed (aidesth¯eis), my child’206 – invites him to transcend the very immaturity it acknowledges. His shyness subdued, Chaereas now serves to enact his adulthood through narrative. Hermocrates helpfully fills in the initial part of the story, allowing the young man to focus upon his own exploits: sailing, enslavement, crucifixion and, finally, the rebellion that proves his manhood – a story that, as recent scholarship has noted, does not fully match up with the primary narrator’s account.207 The story culminates in a series of sentences featuring first-person verbs expressing his decisiveness, vigour and (above all) power: ‘I achieved great deeds . . . I personally subdued Tyre, hard to capture though it is . . . I was chosen as captain, fought a naval battle against the Great King, and became master (kurios) of Aradus . . . I had it in my power (edunam¯en) to make the Egyptian the lord of all Asia’.208 The empowered actor of this narrative is, of course, also its newly empowered author. These narrative acts, then, serve to fix the meaning of the stories, in particular as expressions of the social identity of their actors/authors. In 203 204

205 206 208

Smith (2007) 220–5, 231–2. Hom. Od. 2.40–79. The Odyssean colouring of Chariton’s scene is enhanced by the distinctive Homeric phrase ‘starting from the point where’ (›nqen —lÛn, 8.7.9 ∼ Hom. Od. 8.500, 14.74). Chariton also uses this phrase at 1.7.6, 5.7.10. ßknei Cair”av, Þv ‹n –pª pollo±v tän oÉ kat‡ gnÛmhn sumb†ntwn a«doÅmenov, Char. 8.7.4. mhd•n a«desqiv, å t”knon, Char. 8.7.4. 207 de Temmerman (2009a) 258–60. ›rga meg†la dieprax†mhn [cf. ›rgon . . . m”ga of the capture of Tyre at 7.2.6] . . . TÅron dus†lwton oÔsan –ceirws†mhn aÉt¼v kaª naÅarcov ˆpodeicqeªv katenaum†chsa t¼n m”gan basil”a kaª %r†dou kÅriov –gen»mhn . . . –dÅn†mhn . . . kaª t¼n A«gÅption ˆpode©xai p†shv tv %s©av desp»thn, Char. 8.8.9–10. Chaereas magnifies his account with not only Herodotean echoes (the repeated use of ˆpode©knumi and the reference to –rga meg†la alludes to the preface of the Histories, no doubt also underlining the theme of military conflict between East and West), but also tactical assimilations of his own triumphs with those of Hermocrates, his primary addressee (compare katenaum†chsa with Hermocrates’ defeat of the Athenians: katenaum†chsav, 1.11.2).

First romances

65

each case, after an initial aporia generated by embarrassment (aid¯os, the Greek superego), the speaker takes control of the story and fashions it into a statement of identity: Callirhoe’s as a faithful wife, Chaereas’ as a citizen–warrior worthy of the hand of Hermocrates’ daughter. We can see, then, the Fishy temporality inherent in meaning-making: for narratives to signify they need to be worked through, shaped and mastered. Yet for all that these character-bound narrations attempt to impose a final and total meaning on and control over the narrative, Chariton also dramatises the limitations of that process. Chaereas does not have total control over his narration, at least not at the start: Chariton goes out of his way to emphasise the crowd’s role in shaping the story that is told. Initially, Chaereas ‘began at the end, not wishing to upset the people with the initial (pr¯otois), gloomy (skuthr¯opois) events’.209 This inversion of narrative order is at one level an amusing rewriting of Odysseus’ words when he begins his account of suffering and manly endurance (‘what shall I say first, what last?’).210 It also picks up the narrator’s metadiegetic comments at the beginning of book 8, which we have already considered, to the effect that this final book will be ‘most pleasurable’, and will serve as a ‘cleansing of the gloomy (skuthr¯op¯on) events in the initial (pr¯otois) books’.211 With this echo of both Odysseus and Chariton’s primary narrator, Chaereas is at one level transformed into a narrator. At another level, however, this shows his immaturity in this role. A narrative consists not just in the happy ending, but in the totality of the account. ‘We beg you’, counter the people, ‘start from the beginning, tell us everything, leave nothing out’.212 The shy Chaereas is being coerced into publicly coming to terms with the entirety of his story, even the parts that shame him. As so often in imperial culture,213 oral performance is rerouted and, finally, determined, by audience response. The act of narration, then, is a dynamic, evolving process, not simply an instantaneous and unilateral imposition of the narrator’s fully formed will. This act of narration, this dramatisation of the creation of the ‘meaning’ of Chariton’s story, is irrevocably linked to closure. Chaereas should, Hermocrates replies, cast aside his embarrassment, ‘even if you tell us something very grievous or very bitter; for the luminous ending that has taken 209 210 211

212 213

ˆp¼ tän teleuta©wn ¢rxato, lupe±n oÉ q”lwn to±v prÛtoiv kaª skuqrwpo±v t¼n la»n, Char. 8.7.3. This passage is discussed more fully in chapter 5, below. t© prät»n toi ›peita, t© d ì Ëst†tion katal”xw; Hom. Od. 9.14. kaq†rsion . . . tän –n to±v prÛtoiv skuqrwpän, Char. 8.1.4. On the Aristotelian echo, see Rijksbaron (1984), correcting M¨uller (1976) (who believes that Chariton has failed to understand Aristotle). –rwtämen, Šnwqen Šrxai, p†nta ¡m±n l”ge, mhd•n paral©phiv, 8.7.3. Korenjak (2000).

66

Part i Returning romance

place throws all the earlier events into the shade’.214 Two points call for particular emphasis. First, Hermocrates asserts that the story is over, its completeness boundary-marked by the ‘ending (telos) that has taken place’. Narrativisation marks the finalisation of narrativity, as the narrative catches up with its self-begetting. Second, awareness of the end radically changes the entire perception of the story, converting a narrative of dark suffering into one of light. This is perhaps the most profoundly self-reflexive moment in Callirhoe, signalling as it does a feature so constant in romantic narrative as to be practically constitutive of it: the tension between the characters’ miserable experience of events mid-plot, as the narrative unfurls, and the revisionist sublimation of that experience in the happy ending. In a weak sense, this is true of any narrative, which will both naturalistically mimic the indeterminacy of life as it is lived and play on the ‘preplottedness’ of the authorial plan. ‘The function of a work of art as a finite model of a “speech text” of real facts which is by nature infinite makes the factor of delimitation, of finiteness, the necessary condition of any artistic text in its primary forms.’215 The Greek adventure romance, however, exploits this property of all narrative to the full, widening the gap between the (actorial) experience of infinity and the (narratorial) awareness of delimitation to an unparalleled degree. The final fly in the closural ointment is the way in which, in both Xenophon and Chariton, the romantic narrative overspills the final act of narration. This is subtler in Xenophon, where the narrative continues for three sentences after the lovers’ dedication in the temple of the ‘graph¯e [painting/story] of all that they had endured and all that they had done’. What is interesting about this narrative overspill is that it focuses upon characters other than Anthia and Habrocomes: their parents (who have died of grief ), the slaves Leucon and Rhode, and the erstwhile bandit and bereaved pederast Hippothous. This subversive tactic reminds us insistently that narrative is always presented from a certain point of view: the ‘official’ story of Anthia and Habrocomes in the temple does not capture the variety of experiences undergone by all the figures. Their stories have different rhythms and temporalities, and not everyone gets their happy ending (a point that we shall pick up in chapter 4). In Callirhoe, similarly, the sanctioned, official version of events is offset by an alternative one. While Chaereas is orating in the theatre, Callirhoe 214

215

kŠn l”ghiv ti luphr»teron ¢ pikr»teron ¡m±ná t¼ g‡r t”lov lampr¼n gen»menon –piskote± to±v prot”roiv Œpasi, Char. 8.7.4. Note that Hermocrates’ t¼ . . . t”lov . . . gen»menon picks up the narrator’s initial p†qov –rwtik¼n . . . gen»menon, 1.1.1, an elegant circularity. Lotman (1976) 10.

First romances

67

visits Aphrodite’s temple, gives thanks for her and Chaereas’ safe return home, and prays for a happy future (8.8.15–16). This final act of private reverence contrasts visibly with Chaereas’ public performance, reminding us that this is not, in fact, his story. The narratorial sphragis that closes the text insists as much: ‘this is the extent of my story about Callirhoe’. This sentence, harking back to the title,216 reminds us that despite Chaereas’ public monopolisation of the narration, the narrative actually belongs to Callirhoe. What is more, we know that Chaereas does not know everything: Callirhoe has secretly written to Dionysius: ‘this is the only thing she did unbeknown to Chaereas; she took every effort to keep it secret, knowing of his innate jealousy’.217 (And the reference at this point to her son may well remind us of another deception she has practised, namely telling Dionysius that he is the father.) Her entrusting of her and Chaereas’ son to Dionysius to rear is a significant loose end, an unarticulated narrative ‘aftermath’ that problematises the romance’s closural cadence.218 The romantic ending, then, certainly does enshrine the identity of its protagonists as newly matured, ideal agents within the polity. But it does more than this. Identity is here encoded in narrative, and narrative is a slippery thing: the first-century romancers show great awareness of the constructedness, the partiality, the limitations of narrative, and hence also of the identities that they create. The first-century romances are products of a time of cautious hope in the Greek cities of Asia Minor, a tentative belief that communities can be reborn after struggles and sufferings. This is particularly visible in Chariton, where the theme of post-conflict renaissance is strong. But compromises need to be made, accommodations need to be reached, particularly with foreign powers: here represented in the guise of Persia (but readable as a figure for Rome), transmuted from its traditional status as entirely hostile other. Persia can now be dealt with. Underlying all the idealism of the first-century romance is an awareness of the pragmatism that sustains the myth of a happy community. At the deepest level, then, Chariton and Xenophon both, ultimately, represent optimistic visions of the renewability of the urban community. 216 217

218

Whitmarsh (2005b) 590. toÓto m»non –po©hse d©ca Cair”ouá e«du±a g‡r aÉtoÓ tŸn ›mfuton zhlotup©an –spoÅdaze laqe±n, Char. 8.4.4. On such narrative aporiai in Chariton’s last book, see esp. Brethes (2007a) 179–82. There may even be more to the text than we can know. In history, Hermocrates (the father of the romantic Callirhoe) was succeeded by a famous Syracusan tyrant, Dionysius I; and ancient sources tell us that he married a daughter of Hermocrates, whom he brutally mistreated (Plut. Dio 3; Diod. Sic. 13.112). See further Naber (1901) 98; Perry (1967) 138–9; Hunter (1994) 1056–7. On narrative ‘aftermaths’ see Roberts (1997).

68

Part i Returning romance

They are best understood as textualised transcriptions of the civic festivals and communal activities that they both describe: these romances seek to capture the sense of joyous celebration of the permanence of the city that festival culture enacts ritually. But this textualisation is itself significant. The first-century romances are not functionally equivalent to ritual practice; they are artful, self-conscious narratives, and with narrative comes the depth, richness and complexity that we have traced in this chapter. In the following chapters, we shall see how this sophistication and artistry is exploited in later generations.

chapter 2

Transforming romance Achilles Tatius and Longus

The first-century romances of Chariton and Xenophon mimic the Hellenocentric model of the classic passage-rite myth, whereby the urban, aristocratic Hellenic ‘home’ is offset against the barbarian ‘abroad’. From the second century, romance patterns shift radically. Of the later romances, at least judging by the extant texts, none is centrist in this way. Achilles Tatius’ Leucippe and Clitophon and Longus’ Daphnis and Chloe, the secondcentury1 subjects of this chapter, focus on the Syrian littoral and rural Lesbos respectively, while Heliodorus’ fourth-century Charicleia and Theagenes (the subject of the next) ends in Ethiopian Meroe. Fragmentary texts are, in the nature of things, harder to interpret (and indeed date) with confidence, but the two remaining romances for which the plot is relatively clear fit this trend away from first-century urban Hellenocentrism. The first-century Metiochus and Parthenope, which survives in Greek fragments and a Persian version, is centred on the court of the tyrant Polycrates of Samos, while the second-century Babylonian affairs of Iamblichus is the only known romance entirely to avoid Greek figures and Greek settings. It looks very much, then, as if the romance form shifted radically at the end of the first century, claiming as its own the margins of the world rather than its centres. It is possible, of course, that this pattern is nothing more than a coincidence, the result of nothing more than the aleatory processes of transmission. The texts that we call Ninus and Sesonchosis, the papyrus fragments of which probably date to the first century, were certainly set in non-Greek environs and among non-Greek communities, and may have incorporated return narratives. But it is just as likely that they did not, and that they belong to the broader family of ancient novels written in Greek (a family that includes, for example, Joseph and Aseneth, Antonius Diogenes’ Wonders beyond Thule (which survives in fragments 1

Longus’ date is hypothetical, but likely to be second- or early third-century. See appendix on dates.

69

70

Part i Returning romance

and summary)2 and the Journal of ‘Dictys of Crete’ (the Latin form of which survives complete, the Greek only in fragments)).3 What is more – and this is crucial – the trajectory away from urban Hellenocentrism is only one of a series of interrelated tendencies in the later romances. The first is the relocation of the marriage from the beginning to the end of the narrative. As a result, the return is seen not simply to restore a prior state of affairs, but to transform the identities of the couple. This, I argue, is the most obvious token of a general amplification of the transformative power of romance narrative. We saw in the previous chapter that romance always supplies a centrifugal, transformative push, tensed against a centripetal pull back towards the re-establishment of preexistent norms. The later romances place considerably more emphasis on the transformative vector. The second phenomenon is the increased emphasis on the textualisation of narrative. The first-century romances already (as we have also seen) show considerable awareness of the constructedness of narrative, of the fundamental difference between the plots that human authors shape (Erz¨ahlung) and the experience of life as it is lived (Erlebnis). In the later romances, this play between artifice and nature is exploited much more heavily. For example, the graph¯e (painting/narrative) describing the experiences of Anthia and Habrocomes at the end of Xenophon’s romance becomes, in both Achilles and Longus, an elaborate ecphrasis and a pretext for the very narrative itself. Chaereas’ oral summary of his feats at the end of Chariton is transformed, in Achilles Tatius, into a first-person narrative that occupies almost the entirety of the text. The effect of these processes is to drive a firm wedge between narration and experience, levering open a large space for irony and play, and also focusing even more attention on narrative as plastic, constructed form. At a time when literature and the plastic arts were placing ever more emphasis upon the non-natural, fictive power of artistic mimesis, romance narrative was itself shifting the attention from the message to the medium, from the choses to the mots.4 As Edward Said 2

3 4

Since the publication of SW, P.Oxy. 4760–1 have been published: 4760 is almost certainly from Antonius, 4761 probably. I am not convinced by suggestions that he may be first century (e.g. Bowie (2007)), and faute de mieux prefer a second-century date; but the evidence is too slight to allow any definite conclusion. See also Morgan (1985) 487–9 against Photius’ suggestion (Bibl. cod. 166 111b–12a) that Antonius is likely to have been the earliest of the novels, which is probably based on a naively literal reading of the pseudo-documentarist device. The long-recognised fragments of the Greek text (P.Tebt. 268; P.Oxy 2539) will soon be supplemented by the publication of further Oxyrhynchus papyri (Dirk Obbink reports). For mimesis in imperial Greek culture see Whitmarsh (2001a) 41–89. The final allusion is to Foucault (1970), of which the French title was Les mots et les choses: the final section posits a shift in the understanding of discourse during the period of the enlightenment, when language was (he claims) held to have lost its claim to transparency.

Transforming romance

71

has argued, such eras will see a greater emphasis on narrative as process, as vehicle or as material substance, rather than as a pellucid window onto the world.5 contexts How can we explain these changes in historical terms? Here, as elsewhere in this book, I wish to avoid historical determinism, but, even so, there are phenomena that help to contextualise these trends. The most significant is the emergent interest in the malleability and transferability of identity.6 The steady trickle of elite Greeks attaining Roman citizenship and entering the Roman senate that began in the late republic became much more substantial,7 until, finally, in 212 ce Caracalla’s Constitutio Antoniniana Romanised the entire free population of the empire. Conversely, we see a greater focus on the Hellenisation of non-Greeks. Although Roman semiotic Hellenism stretched back to at least the second century bce (and arguably well beyond that),8 the second century ce saw a step change, in that emperors were now explicitly self-fashioning as Greek. Nero and Domitian, like Antony before them, had adopted some markers of Hellenic stylisation, but it was Hadrian (117–38) in particular who cultivated an image of himself in unmistakeably Greek terms: accompanying the famous beard,9 boyfriend and poetry, concrete symbols of his affiliation, were a series of practical policies supporting Greece: the patronising of Greek literary talents,10 the renewal of urban landscapes and the foundation of new cities,11 and the creation of the famous, if still partially mysterious, institution called the Panhellenion, a league (based on the Delphic Amphictiony) of cities who could prove their Greekness genealogically, centred in Athens and built around the cult of the emperor.12 From here on until the so-called ‘crisis’ of the mid-third century, all emperors would present themselves in Greek guise. Even more important for the romancers is the emphasis in the literature upon the Hellenisation of easterners, which, while of course in reality it reflects a process that was as old as archaic colonisation, achieves in this

5 7 8 9 11 12

6 See more fully Whitmarsh (2001a). Said (1983) 101, on Conrad. See esp. Halfmann (1979) on eastern senators, and further literature cited at Whitmarsh (2001a) 18. Whitmarsh (2010a), with further literature. Vout (2006) emphasises the iconographic polysemy of the beard. 10 Fein (1994). On Hadrian’s provincial building, see Boatwright (2000), emphasising that Hadrian’s civic munificence also extended to central Italy and Africa. Discussion and further literature at Romeo (2002).

72

Part i Returning romance

era an unparalleled visibility and salience:13 well-known examples include Philostratus’ Ninevan Damis who articulates his desire to ‘become Greek’ by associating with the great philosopher Apollonius,14 the Syrian Lucian’s self-fashioning as a Hellenised barbarian, and the Gaul Favorinus’ deconstruction of the opposition between true and self-made Greeks.15 Greek identity thus became both significantly more attractive and significantly more attainable in the second century. This privileging of Greekness, however, also made it a target. The early imperial period saw numerous claims (in Greek) to the effect that Greek ideas had been pre-empted by Babylonians, Jews, Egyptians and others. Again, the proto-Bernalian claim itself is not new: it can be glimpsed already in the fourth century (Plato’s Egyptian priest teases the Greeks as ‘children’ at Timaeus 21b), and is widespread in Hellenistic historiography: for example, in the Jewish historian Eupolemus claimed that Moses was the ‘first’ wise man, the inventor of the alphabet, which passed to the Greeks via the Phoenicians (FGrH 723 F1); or in Hecataeus of Abdera, who makes Egypt the origin of geometry, astronomy and arithmetic (264 F1). For all the deep roots of this tradition, the imperial period saw an intensification of such claims, notably in Josephus’ Jewish antiquities and Against Apion (late first century ce), and continuing forcibly in the second-century Christian tradition: Justin Martyr, Tatian, Clement and Origen represent a strong collective statement that Mosaic law precedes, influences, and morally trumps Greek philosophy.16 It is, perhaps, particularly this contemporary pressure from Christians (more directly than the non-Christian Hellenistic precursors whom he explicitly identifies) that inspires Diogenes Laertius, in (probably) the early third century ce, to begin his Lives of the eminent philosophers with an attack on those who claim that the philosophy is a barbarian invention (1.3).17 The crucial point, however, is that this entire debate is conducted in Greek itself. Hellenism might be opposed, but it is inescapable. The second-century romance reflects this fluid world, transforming the Hellenocentric paradigms encoded in the first-century texts by placing 13

14 15

16 17

Whitmarsh (2001a) 90–130. The idea that barbarians can become Greek through education is, assuredly, as old as Isocrates (esp. Panegyr. 50). More generally on paideia in the imperial period see Swain (1996) 17–100; Schmitz (1997). ï Ellhn gen»menov, 3.43. Gleason (1995) 16–17 and Whitmarsh (2001a) 119–21, 169–78 on Favorinus, with further literature (adding K¨onig (2001) more generally on the context of the speech; Amato (2005) offers an excellent new edition). On Lucian’s Hellenism see Swain (1996) 298–329 and Whitmarsh (2001a) 247–94. The tradition (with its Hellenistic background) is economically surveyed by Droge (1989); for interpretation, see esp. Stroumsa (1999) 57–84. As suggested at Whitmarsh (2007a) 38–9.

Transforming romance

73

greater emphasis upon alterity and the mutability of identity, and by focusing attention self-reflexively upon the very narrative processes of selfconstruction. hellenism decentred Whereas Chariton and Xenophon construct the Greek city as the sine qua non of civilised living, second-century romances place much less emphasis upon civic identity. In Leucippe and Clitophon, the narrator Clitophon announces that his ‘homeland (patris) is Tyre’ (1.3.1), but describes nothing of its people, architecture, religion or institutions. It is not that the romance shows no awareness of institutional structures – the cult temple of Astarte at Sidon (1.1.2), the public buildings of Alexandria (5.1–2) and the courtroom (7.7–13) and temple of Artemis (7.16.2–8.) at Ephesus are all sketched – but the separation–reincorporation narrative is domestic rather than civic.18 The setting for the first two books is Clitophon’s family home (described in lavish detail), and it is their relationships with their father and mother respectively that Clitophon’s and Leucippe’s love for each other compromises (1.11.3, 2.30; cf. 5.18.4). In Daphnis and Chloe, meanwhile, we are certainly told that the protagonists begin life in the polis and end up there, but the narrative shows no interest in their life there. After the proem, the narrative opens with an ecphrasis of Mitylene, and indeed the first word is polis; but the scene immediately shifts 200 stades away, as if programmatically to emphasise the romance’s shift from a civic to a rural focus (1.1). Like Dio Chrysostom’s Euboean oration and Philostratus’ Heroicus, Daphnis and Chloe challenges its readers’ sense that civic living is more civilised, by presenting nature as generally benign and urban wealth as generally corrupt and oppressive.19 The countryside is where they live from early childhood, and is the formative influence upon them, to the extent that, even after the recognition, they continue to live a pastoral existence. They insist on marrying in the countryside, ‘since they could not stand spending time in the city’;20 and afterwards, ‘they led their lives in the pastoral way most of the time’, cultivating the rustic deities, and suckling their child on a goat and a ewe.21 18 19 20 21

See Whitmarsh (2010b). Discussion and bibliography at Whitmarsh (2001a) 100–8; see further below, pp. 96–9. mŸ f”rontev tŸn –n Šstei diatribžn, 4.37.1. t¼n ple±ston cr»non poimenik¼n e²con, 4.39.1.

74

Part i Returning romance

The devaluation of the polis signals in part a greater concentration upon the complex psychology already latent in the first-century romances. But it also marks a cultural repositioning. The second-century texts no longer seek to naturalise a view of the world centred on traditional Greek values; rather, their tendency is to disrupt received constructions of the world, and multiply the cultural perspectives. Iamblichus’ Babylonian affairs takes place entirely in Mesopotamia, and seemingly includes not a single Greek figure. Leucippe and Clitophon begins and ends on the Phoenician littoral. Heliodorus’ Charicleia and Theagenes, the subject of the following chapter, takes us to Ethiopia, for Greeks one of the edges of the earth. Fragmentary narratives also used non-Greek locations, such as Lollianus’ Phoenician affairs (the setting for the surviving fragments is unspecified, but the title is secure) and the story that we call Calligone (set in southern Russia).22 Non-Greek settings per se, for sure, do not begin in the second century, as we noted at the outset of the chapter. As Martin Braun already argued in the 1930s,23 the earliest novels were variants of the ‘national romances’ of the peoples subjected to Greek and then Roman hegemony. Prior to the second century, however, romances on non-Greek themes seem to have focused on iconic individuals of history, particularly rulers. These novels may have incorporated romantic elements (like the Alexander romance’s erotic interlude involving Candace, queen of Meroe: 3.18–23). The second-century romance, by contrast, takes the adventure romance format of Chariton and Xenophon – the marriage theme, the centre–periphery structure, the emphasis on maturation and the role of marriage, and the use of (wholly or largely) invented figures – and introduces into it a much greater sense of cultural plurality. Given the ethnocentrism inherent in the first-century centre–periphery paradigm, this in itself constitutes an implicit challenge to Hellenocentric ways of thinking. This shift in emphasis seems to reflect a diversification of origins among the romancers themselves, although (as ever) our biographical information is sparse. Achilles Tatius is associated with Alexandria by the manuscript tradition and by the Suda.24 It is possible that the association derives simply from the lavish praise of Alexandria found at the outset of book 5, 22

23 24

At a greater distance from the erotic romance, we could also point to Antonius Diogenes’ Wonders beyond Thule and Lucian’s True stories, although Hellenistic precedents certainly existed for this kind of fantastic travel narrative (e.g. Euhemerus, Iambulus, Eudoxus, Pytheas). Braun (1934), (1938). Suda s.v. ìAcilleÆv St†tiov (sic), and the MSS titles cited at e.g. Vilborg (1955) 1.

Transforming romance

75

but, in general, scholars have found enough convincing evidence of (apparently) first-hand knowledge of Egypt.25 Some have claimed that the surname ‘Tatius’ (T†tiov) derives from the Egyptian god Thoth (∼ Greek Tat),26 which would be an excellent sobriquet for an author, given that Thoth was in some traditions the inventor of writing or ‘father of letters’.27 Others, however, have seen it simply as a form of the Roman name Tatius, common enough in the Greek world.28 On balance, it is likely (albeit far from certain) that Achilles was Alexandrian. Perhaps he considered himself culturally Greek but ‘Egyptian by race’, like his own, equally Iliadically named, character Menelaus.29 Meanwhile Iamblichus was (according to a scholion on Photius, which seems to derive from the romance itself ) a native Syrian for whom Greek was the third language, learned after ¯ Babylonian.30 His name is a Hellenised form of the Semitic YMLK ’EL (=‘El rules’, or alternatively ‘El makes him king’), relatively common in the areas of Syrian Emesa and Palmyra.31 Heliodorus, as we shall see in the following chapter, also seems to have hailed from Emesa (10.41.4). (As to Longus, all is speculation: even the name, in Greek Loggos, may simply be a corruption for logos, ‘story’).32 These later romances display a much greater interest in, and narrative exploitation of, cultural heterogeneity. In Chariton and Xenophon, as we saw in the previous chapter, non-Greeks unproblematically speak Greek, recognise Greek institutions, and worship Greek gods. Leucippe and Clitophon, by contrast, thematises difference from the start. The romance begins with a brief geophysical description of Sidon: 25 27 29 30

31 32

Rommel (1923) 78–81. 26 E.g. Vilborg (1962) 7. p†thr gramm†twn, Pl. Phaedr. 275a. 28 E.g. Plepelits (1980) 2. t¼ . . . g”nov A«gÅptiov, 2.33.2; cf. 3.19.1, with n. 54 below, and –moª Foin©kh g”nov at 1.3.1 (Clitophon). This passage is cited at Habrich (1960) 2 and translated at SW 181. It conflicts in points of detail with Photius’ own summary (at Bibl. 75b 27–41, cited Habrich (1960) 32–4; translated SW 194), notably in that it makes the author a Syrian rather than a Babylonian. The scholarly consensus, that the scholion is correcting Photius’ misreading, is probably right: see esp. Millar (1993) 489–92, SW 181–2. Iamblichus is also called ‘Syrian’ at Theodorus Priscianus Eupor. 133.5–12 (Rose). It is worth noting that elsewhere (Bibl. cod. 181 = 125b) Photius claims that an Iamblichus (perhaps ours, but not necessarily) was from Emesa (Heliodorus’ homeland: see the following chapter), and descended from an ancient dynasty. Chad (1972) 143–4. Scholars have made much of the fact that the name is attested epigraphically for Lesbos; this, coupled with an apparently accurate knowledge of the island (Mason (1979), (1995); Green (1982)), makes a Lesbian origin possible. But the name is common enough: found once on Crete and four times in Cyrenaica (LGPN 1:289); four times in Athens (LGPN 2: 285); and twice in Thrace (LGPN 4: 211).

76

Part i Returning romance

Sidon is a city on the sea. The sea belongs to the Assyrians, the city is the Phoenicians’ mother-city, and its people fathered the Thebans. In the folds of a bay lies a twin harbour, broad and gently enclosing the sea: where the bay bellies out down the flank of the coast on the right, another mouth has been carved out, where the water flows back in. Thus a second harbour is born from the first, so that trading vessels can winter there in the calm, while they can pass the summer in the outer part of the bay.33

The first three proper names are exotic: Sidon, the Assyrians, the Phoenicians. The Thebans, meanwhile, play a double role. On the one hand, the allusion to a familiar Greek people implies a Greek framework of reference, controlling and taming the Phoenician otherness (rather as Philostratus’ narrator in Apollonius describes the Indian landscape by comparing it to canonical Greek sights such as the Athenian acropolis: VA 2.20, 2.23, 2.27, 3.13). Intercultural description, as Edward Said stresses, creates cultural subjectivity: it positions ‘us’ as the controllers of knowledge, and the other as its object.34 On the other hand, the description of Thebes as a colony of Sidon implicitly places the Greek polis in a relationship of dependence and secondariness – especially in the wider context of the ‘culture wars’ of Achilles’ time, when cultural priority was such an issue. The effect is disorientating and confusing.35 This sense of disorientation is exacerbated by the bizarre syntax of the Greek: there are no verbs or connective particles at all until we reach the geophysical description, and even there we are thirty-three words into the romance before we reach a main verb (koilainetai, ‘bellies out’). The phrasing, moreover, seems calculated to confuse, particularly in the third and fourth sentences, which literally read: ‘mother of Phoenicians the city; of Thebans the people father’. Each sentence consists of three elements – an unarticulated nominative noun denoting a parent, an unarticulated genitive plural noun denoting a people, and a nominative articulated noun (‘the city’, ‘the people’) – but these elements occupy different positions. The rendering given above36 is the result of much labouring. In fact, a more natural way to construe the Greek syntax is: ‘the mother of the Phoenicians 33

34 35

36

SidÜn –pª qal†tthi p»livá %ssur©wn ¡ q†lassaá mžthr Foin©kwn ¡ p»livá Qhba©wn ¾ dmov patžr. d©dumov limŸn –n k»lp platÅv,  r”ma kle©wn t¼ p”lagov· ¨‚ g‡r ¾ k»lpov kat‡ pleur‡n –pª dexi‡ koila©netai, st»ma deÅteron ½rÛruktai, kaª t¼ Ìdwr aÔqiv e«sre±, kaª g©netai toÓ lim”nov Šllov limžn, Þv ceim†zein m•n taÅt t‡v ¾lk†dav –n galžn, qer©zein d• toÓ lim”nov e«v t¼ prok»lpion, 1.1. Said (1978). Even more so when we consider that Cadmus is elsewhere said to come from Tyre, not Sidon (Vilborg (1962) 18). Has Achilles, or Clitophon, made a mistake? Or is this deliberate? If so, to what end? Again, Achilles’ primary aim seems to be discombobulation. Adapted from Whitmarsh (2001b).

Transforming romance

77

is the city; of the Thebans, the people is the father’ – albeit this makes less good sense. A further challenge is presented by the play between ‘mother’ and ‘father’ here? What is at stake in the gender differentiation? At one level, the answer lies in a play upon Greek grammar, since ‘city’ is feminine and ‘people’ is masculine. But that answer just raises another question: why is it that the city is the parent of the Phoenicians, and the people that of the Thebans? What is more, referring to a city as a ‘mother’ seems to imply a colonial mother-city (m¯etropolis), and that description would seem to be more appropriate to the relationship between Sidon and Thebes. Readers of the first-century romance would have been bamboozled. Both Chariton and Xenophon begin with unproblematic descriptions of their central figures and their homelands (Chariton having introduced himself beforehand): ‘Hermocrates the general of the Syracusans, the victor over the Athenians, had a daughter called Callirhoe’; ‘there once was a citizen of Ephesus called Lycomedes, one of the most powerful men in the city.’37 Achilles’ perplexing beginning, relative to his predecessors, does not simply problematise interpretation; it also highlights the issue of subjectivity, of the vantage from which the scene is being observed: we are struggling to construe the scene not just because the text is challenging and experimental, but also because what is being described is alien to our experience. Description loses its easy naturalism, its pretensions to pellucidity, and becomes instead an object lesson in the perils, and the politics, of intercultural interpretation. culturally speaking Another factor that militates against transparency is the characteristic emphasis upon the ways in which narrative is mediated. I have been writing as though the cultural distance we are considering were the space between the reader and Sidon. But interposed between the two, of course, stands the narrator’s own interpretation: we readers interpret the narrator interpretating Sidon. After the passage quoted and discussed above, we encounter the figure of this mediator: ‘It was there that I arrived, a survivor of a severe storm, and made my thank-offerings for my rescue to the 37

ëErmokr†thv ¾ Surakos©wn strathg»v, oÕtov ¾ nikžsav ìAqhna©ouv, e²ce qugat”ra Kallirožn, Char. 1.1.1; §n –n ìEf”swi ˆnŸr tän t‡ präta –ke± dunam”nwn, Lukomždhv Ànoma . . . , Xen. Eph. 1.1.1. Longus and Achilles supply contextualising material both in their external frames and in the secondary narration that thereafter becomes the principle focus: cf. Long. 1 praef. 1 (–n L”sbwi qhrän –n Šlsei Numfän q”ama e²don k†lliston æn e²don), 1.1.1 (p»liv –stª tv L”sbou Mitulžnh . . . ); Ach. Tat. 1.1.1 (SidÜn –pª qal†tthi p»liv), 1.3.1 (–moª Foin©kh g”nov, TÅrov ¡ patr©v, ktl.). Prefatory strategies are discussed by Morgan (2001).

78

Part i Returning romance

Phoenicians’ goddess, whom the Sidonians call Astarte.’38 Who is this ‘I’? Scholars conventionally refer to an ‘unnamed narrator’; I suspect, however, readers without the benefit of Genettian narratological categories39 would have been more disposed to identify a narrative ‘I’ more or less directly with the author. I do not mean that the shipwreck story asks to be taken as autobiographically true, but that ancient readers were likely to have associated the claim (however fictitious they may have seen it as) with the author’s own voice – as, for example, Augustine famously reads the ‘I’ of Apuleius’ Metamorphoses as an autobiography of Apuleius himself, ‘whether he recorded it or invented it’.40 I shall argue presently that the link between the narrator and Achilles enriches the romance. My point for now, however, is simply that the use of this technique lays heavy emphasis upon the question of who is speaking: not the diaphanous, authoritative narrator of Chariton and Xenophon, rather a personalised but mysterious individual, frail enough to be shipwrecked; someone with his own story, his own identity, and (by implication) his own particular set of narrative filters. Indeed, we can perhaps see these filters at work in the opening passage cited above. The one thing that we do learn about the narrator’s background, presently, is that he is er¯otikos (1.2.1): ‘in love’, or perhaps ‘of amorous disposition’, or even ‘an erotic expert’.41 Perhaps this 38

39 40

41

–ntaÓqa ¤kwn –k polloÓ ceimänov, sästra ›quon –mautoÓ ti tän Foin©kwn qeiá %st†rthn aÉtŸn kaloÓsin o¬ SidÛnioi, 1.1.2. Most MSS omit qei, which appears only in F, described by Vilborg as deriving from a scribe who ‘used his own judgement in interpreting and emending the text’ ((1955) lxx). Vilborg himself prints qei, despite suggesting that it is an interpolation. Part of the reason for suspecting qei is the claim that elsewhere Achilles uses qe»v of goddesses ((1962) 19). This is not true: qe† is in fact attested by V, G and F at 2.36.2, a reading that Vilborg himself, like all modern editors, adopts. The erroneous claim is repeated by Diggle (1972), who wants to read ti tän Foin©kwn %frod©thi – unjustifiably interventionist, in my view. I have maintained the order of VGE (followed by Vilborg, but not Garnaud) for the second sentence quoted here. ‘That it is essential not to confuse author and narrator has become a commonplace of literary theory’ (Chatman (1978) 147). aut indicavit aut finxit, Civ. 18.18 – a confusion that is of course already seeded in Apuleius’ Metamorphoses itself, where Lucius claims to hail from Madaurus, Apuleius’ hometown (11.27; and see, among the many studies on this phenomenon, esp. Laird (1990) on what I shall later in this chapter call the metaleptic quality of such passages). ‘Augustine clearly takes it as an autobiography, whether real or fictitious: for although he denies the possibility of metamorphosis and doubts the sincerity of Apuleius’ account, he assumes without question that Apuleius is claiming to relate his own experience – that he is the Lucius of his novel. The assumption continued to be unquestioned for at least a thousand years, and the identity of Apuleius and Lucius was to play a major role in the interpretation of the Golden ass’ (Gaisser (2008) 33). Similarly, Photius conflates narrator and author in the Metamorphoses of Lucius of Patrae: Bibl. cod. 129, with Whitmarsh (2010c). Such confusion is not really straightforward naivety – neither Augustine nor Photius believes that the events described actually happened to the author – but rather the result of a different intellectual mindset, in which first-person statements are attached much more closely to their speakers. This last interpretation I owe to Ian Repath, whose rich and suggestive book Playing with Plato will argue (among many other things) that the unnamed narrator is a Socratic figure.

Transforming romance

79

disclosure retrospectively explains the superabundance in the opening passage of terms for parenthood (‘mother’, ‘father’; also perhaps ‘twin’, didymos, and ‘is born’, ginetai), the corporeal topography (kolpos (twice; also prokolpion) is not only ‘bay’ but also ‘bosom’ or ‘lap’; pleura, ‘flank’, can also mean the human flank; koilanetai, ‘bellies out’, literally implies hollowness, suggesting the body’s cavities; stoma usually refers to the human ‘mouth’), and the emphasis upon fluidity: an er¯otikos might find much to conjure with in the phrase ‘another mouth . . . where the liquid flows back in’.42 How culturally inflected is this mediation? Although it is not stated where the narrator comes from, or where he is going to (all that we can say is that he seems to be at Sidon by accident, and apparently as an outsider), the narrator seems to engaged in interpretatio Graeca, the translation of foreign elements into a Greek register: this is strongly implied by the reference to ‘the Phoenicians’ goddess, whom the Sidonians call Astarte’. In an influential discussion, Daniel Selden has argued that the narrator misperceives Sidonian cult from his Greek perspective, whereas Achilles also makes available (to those in the know) a different, Sidonian interpretation; the text thus embodies the rhetorical figure that he names ‘syllepsis’, i.e. an availability to be read in two opposing ways.43 Selden’s argument focuses primarily upon the lengthy ecphrasis of a painting of Zeus’ rape of Europa (1.1.2–13); readers familiar with west-Asian mythology would, he claims, have interpreted this differently, as representing the dominance of Astarte (i.e. the semitic cosmic deity Iˇstar, sometimes identified with the moon) over the sea: ‘The text strategically accommodates both possibilities, so that depending on the reader’s frame of reference, Hellenic or Phoenician, the image can be decoded in two opposing ways.’44 Despite its attractions, there are a number of problems with the argument as formulated. First, the temple of Iˇstar may be Sidon’s most iconic landmark,45 and we might assume that the narrator’s dedication of his thank-offerings takes place there, but the painting is not located (as Selden claims)46 within it: what the text actually says is ‘when I had made my 42 43 44

45 46

st»ma deÅteron . . . kaª t¼ Ìdwr aÔqiv e«sre±. Selden (1994) 50–1, supported with qualifications by Morales (2004) 38–48. Selden (1994) 51. The two opposing modes of interpretation are: (i) the Hellenic, which sees Europa as abducted by the bull; and (ii) the West Semitic, wherein the goddess is in control, leading her mate out to sea. Generally, on Phoenicians in the romances, see Briquel-Chatonnet (1992). Cf. Luc. DDS 4 (¬r¼n . . . m”ga). On the evidence for the temple see further Lightfoot (2003) 297–301. Selden (1994) 50; Morales (2004) 37. Lightfoot (2003) 299 recognises the problem.

80

Part i Returning romance

thank-offerings . . . I undertook a tour of the rest of the city’,47 and it is there that he saw the ‘votive picture’, while browsing ‘the sacred dedications’. It is possible that the text is corrupt – ‘sacred dedications’ are more likely to be found in a temple than in urban space48 – but as it stands the text does not connect the painting directly with Astarte. Second, Selden wants to set the narrator’s identification of the figure as Europa against what he takes as Clitophon’s subsequent correction, when in his later narrative he describes Leucippe as ‘like Selene I once (pote) saw drawn on a bull’.49 For Selden, Clitophon is referring back to the painting seen at the outset, and supplying Selene (=Astarte) as the subject. However, even if we discount the textual difficulties here too,50 it is far from self-evident that he is actually referring to the picture described at the beginning of the text: ‘a picture I once saw’ is not ‘the picture that you and I saw just now’.51 Finally, the parallel Selden draws with the Lucianic On the Syrian goddess (4) actually works against his reading: in the Lucianic text it is the Sidonian priests who identify Astarte with Europa, while the narrator (a non-Sidonian Phoenician) offers Selene. In the Lucianic passage, Europa is the local reading, and Selene the external one – the exact inverse of the configuration Selden wants.52 Selden is broadly right about the unnamed narrator. Although he is not explicitly marked as ‘Greek’, this is implicit: his role (as we have seen) certainly seems to be to translate local, Phoenician phenomena into a register accessible to a panhellenic audience. Clitophon, however, is more problematic for his argument, since – despite what Selden claims – he shows no signs of Semitic awareness. Though ‘Phoenicia provides my ancestry’,53 47 48

49 50

51

52

sästra ›quon –mautoÓ . . . peri·Ün oÔn kaª tŸn Šllhn p»lin, 1.1.2: see below for my proposed emendation. My proposed emendation is tŸn aÉlŸn per©bolon (‘the temple precinct’) for tŸn Šllhn p»lin (‘the rest of the city’). The phrase is not directly paralleled, but for aÉlhv per©bolov cf. Ael. Ar. Rom. 29 (where emendation to aÉl»v (so Oliver (1953); Klein (1983)) is misguided: Pernot (1997) 74 n. 61); Hsch. x 653. It could be, then, that Achilles wrote t¼n aÉlhv per©bolon, although that is palaeographically more difficult. I read this episode as an allusion to ps.-Cebes’ Tabula, where again the painting stands in the precinct (peri·Ün oÔn kaª tŸn aÉlŸn per©bolon kaª periskopän t‡ ˆnaqžmata ¾rä grafŸn ˆnakeim”nhn . . . , Ach. Tat. 1.1.2 ∼ poll‡ m•n kaª Šlla ˆnaqžmata –qewroÓmená ˆn”keito d• kaª p©nax tiv ›mprosqen toÓ neÛ . . . , ps.-Ceb. 1.1.1); cf. also Luc. Tox. 6. Vilborg (1962) 19 notes the high concentration of textual uncertainties in the first page of Achilles. toiaÅthn e²don –gÜ pote –pª taÅrwi gegramm”nhn Selžnhn, 1.4.3. Selžnhn (WMD); EÉrÛphn (VGE). It has been claimed, reasonably, that Selene is the lectio difficilior, since Europa may be a scribal attempt to force the link back to the opening picture (so Vilborg (1962) 21–2). That pote cannot mean ‘just now’ is noted by Vilborg (1962) 20; Lightfoot (2003) 301; Morales (2004) 40. Note also that after beginning his narrative, Clitophon never elsewhere engages the unnamed narrator or refers to the external narratorial circumstances: H¨agg (1971) 125; Morgan (2004d) 495. Lightfoot (2003) 299. 53 –moª Foin©kh g”nov, 1.3.1.

Transforming romance

81

his name and cultural vocabulary are entirely Greek.54 The word ‘barbarian’ in his mouth always carries negative associations.55 At one point, notably, he tells Leucippe the familiar story of Philomela, laying particular emphasis upon the ethical superiority of Greeks over barbarians (5.5.2–3, at 2). His reference points, like those of his cousin Clinias, are entirely drawn from the mainstream Greek tradition; indeed, it is particularly awkward for Selden’s argument that Clitophon never mentions Astarte or any local Phoenician deity. At 1.4.3, the passage discussed in the previous paragraph, he refers to Selene, not Astarte, and indeed elsewhere alludes to Europa on the bull (2.15.4) – precisely the story that (in Selden’s reading) he is supposed to repress. The one exception is the aetiology of the vine at 2.2, where Clitophon supplies a version of the Icarius story, marked as Tyrian in explicit contradistinction to the Athenian: ‘The Tyrians consider that Dionysus is a local god, since they too sing the myth of Cadmus.’56 Even here, however, Clitophon distances himself personally from this story, bookending his account in the Herodotean ethnographic fashion with the phrase ‘so the Tyrians’ story goes’,57 and proceeding to link it to the festival that ‘they’ (third person) hold (2.3.1). Clitophon, then, seems (although we shall qualify this below) ambiguously poised in relation to Tyre: although happy to describe himself as Phoenician by ancestry, he presents himself as aloof from local custom. Now it might be possible to argue that this Hellenised perspective results from the overlaying of the primary narrator’s subjectivity onto an originally Phoenician tale, but this raises a related but arguably greater problem. The text as we have it is multiply embedded: it represents (perhaps) Achilles’ own Alexandrian version of the panhellenic(?) narrator’s version of Clitophon’s Greco-(?)Phoenician version of events. (This narrative Chineseboxing is borrowed from Plato, particularly the Symposium, a vital intertext for Achilles.)58 But, of course, in a work of fiction it is impossible to tease apart these layers, as a historical Quellenforscher might. Neither Clitophon nor the primary narrator exists independently of the text: any account of 54

55 56 57 58

A comparable case is his acquaintance Menelaus, ‘an Egyptian by birth’ (t¼ . . . g”nov A«gÅptiov, 2.33.2; cf. 3.19.1; also above, n. 29), but onomastically Greek, and seemingly differentiated from the ‘Egyptians’ who receive Clitophon’s scorn for their cowardice (4.14.9). See esp. the dismissive or fearful references at 3.9.2. 3.24.3, 4.17.1, 5.5.2, 8.2.3. Kuch (2003) discusses the Greek–barbarian antithesis in the romances. t¼n g‡r Di»nuson TÅrioi nom©zousi —autän, –peª kaª t¼n K†dmou mÓqon Šidousi, 2.2.1. Þv ¾ Tur©wn l»gov, 2.2.6. First noted by Winkler in Reardon ed. (1989) 284 n. 72; more detailed discussion in the forthcoming book of Ian Repath. The general significance of the Symposium for Achilles has been discussed by Morales (2004) 51–3; Laplace (2007), esp. 463–532.

82

Part i Returning romance

the play of voices has to focus solely upon the textual process, and not rely on hypothecations of ‘original’ subjectivities, whether Phoenician, Greek or any other. There are two issues here. The first is the historical question of what Phoenician cultural consciousness might actually have consisted in by Achilles’ time: impossible to answer, given the limited and intractable evidence, but it is highly problematic to use Ugaritic texts from the second millennium as a guide to the thought patterns of Sidonians of the second century ce, who had been creatively hybridising their culture for at least four centuries. The Europa narrative represented a particularly complex case of hybridity: from the third century bce, Phoenicians had appropriated this myth from Greeks and used it as their own narrative.59 Achilles’ Europa story, then, is not a case of Greek misreading of an ‘authentic’ Canaanite myth, but reflects a genuinely Phoenician process of narrative recycling, arguably even a counterhegemonic case of ‘colonial mimicry’ that disturbs and subverts Hellenocentrism by emphasising the Phoenician origin of Greek culture.60 The second issue relates to reading strategies. There is no possibility of neatly separating Achilles’ voices into cultural categories. The play of voices is precisely the point.61 How, then, can we gauge the extent to which Clitophon’s words have been re-encoded as Greek by the primary narrator? At the beginning of Philostratus’ Apollonius, the narrator claims to have ‘rewritten’ (metagrapsai) the crude memoir of Damis, the ‘man of Nineveh’, polishing the style and combining it with other sources. Has a similar process taken place in Leucippe and Clitophon? And if the story has been redacted, to what extent have elements (e.g. Phoenician features) of the ‘original’ been censored out? Has Clitophon been Hellenised by the narrator’s acts of cultural translation? These questions are of course unanswerable: the important point is not to root out demonstrable truths in this fictional text, but to allow these unsettling questions to resonate. Leucippe and Clitophon is a subversive text, and one of the many things that it subverts is a reader’s anticipated confidence in narratorial veracity: in this story, we never quite know whose voice (or voices) we are listening to.62 59 60 61 62

Lightfoot (2003) 297–9. Millar (1983) 48–9 reads Achilles’ account as consistent with this process of hybridisation. For ‘colonial mimicry’ see Bhabha (1994) 85–92, and further below n. 63. Whitmarsh (2003), Morgan (2007a), Marinˇciˇc (2007); see further below. Even if from a strictly narratological perspective there is rarely any ambiguity as to the identity of the narrator: Morgan (2004d) 502–6.

Transforming romance

83

Let us return to the story of Tyrian Icarius, discussed earlier. There, I offered the preliminary claim that Clitophon speaks phrases such as ‘so the Tyrians’ story goes’ (2.2.6). It is true enough that there is no reason to suspect a change of speaker here, even if there is certainly a change of narrative register in this little ethno-mythographic excursus. But might we not attribute this passage instead to Achilles and/or the unnamed narrator, i.e. those narrators concerned with the mediation of discourse between the local and the panhellenic? Or is it perhaps Clitophon himself aping the panhellenising discourse of his Greek interlocutor? Is he engaging in what Homi Bhabha calls colonial mimicry, the ‘ironic compromise’ between the ‘synchronic, panoptical vision of domination’ produced by panhellenic ethnography and the specificity of the local?63 These questions are real in that they are part of the intriguing pleasure of the text; but they are false in that they are irresoluble. In multiply embedded narrative, all voices but the uppermost are evanescent, spectral. The play of voices becomes even more deliciously complex at the beginning of the fifth book of Leucippe and Clitophon, where Clitophon and Leucippe arrive in Alexandria.64 As Stephen Nimis has noted, this episode ‘has the earmarks of a new beginning’.65 The connections with the opening of the romance, indeed, are striking. Book 1 begins with an unnamed narrator arriving by sea at Sidon; in book 5 Leucippe and Clitophon come to Alexandria by boat. Clitophon’s tour of the city (periag¯on . . . emauton, 5.1.5) and its spectacles mirrors the narrator’s initial tour of Sidon (perii¯on, 1.1.2); both figures drink in the lavish sights before them, describing them in erotic terms. What is more, both visit the temple of and pray to a culturally problematic god: Clitophon that of ‘the great god, whom the Greeks call Zeus, the Egyptians Sarapis’;66 the unnamed narrator that of ‘the Phoenicians’ goddess; the Sidonians call her Astarte’;67 and, in both cases, the visit is swiftly followed by the viewing of a painting. More specifically, Clitophon’s eroticised68 description of the city looks back to his first encounters with Leucippe.69 When he comments that ‘the lightninglike beauty of the city immediately confronted me, and weighed down

63 64 66 67 68

Bhabha (1994) 85, paraphrasing Said (1978) 240. This abbreviates the fuller discussion at Whitmarsh (2009b) 44–7. 65 Nimis (1998) 110. toÓ meg†lou qeoÓ, Án D©a m•n í Ellhnev, S”rapin d• kaloÓsin A«gÅptioi, 5.2.1. ti tän Foin©kwn á %st†rthn aÉtŸn kaloÓsin o¬ SidÛnioi, 1.1.2. On the textual issues here see above, n. 38. Morales (2004) 100–6. 69 Giatromanwl†khv (1990) 661–2.

84

Part i Returning romance

(egemisen) my eyes with pleasure (h¯edon¯es)’,70 this combines the effect of Leucippe’s arrival (she ‘struck my eyes like lightning’)71 with the aftermath of the first symposium: the others (we are told) measured their ‘pleasure’ (h¯edon¯en) with their bellies, whereas Clitophon was ‘weighed down with’ (gemistheis) the sight of the girl’s face (1.6.1).72 All of these echoes serve to cast book 5 as a rerun of book 1, with Alexandria replacing Sidon/Tyre. That the description is of Alexandria, of all cities, is especially significant, in that it is (probably) the author’s own homeland.73 If it is right, as I have suggested above, that ancient readers would probably have associated the unnamed narrator of the romance with the author himself, then this passage is truly extraordinary: at the halfway point in the romance, Clitophon visits Achilles’ homeland, just as ‘Achilles’ visits Sidon at the start of the narrative (the two passages, as we have noted, powerfully echoing each other).74 Each is abroad in the other’s home. This makes even more intriguing Clitophon’s curious sentence ‘Many a road criss-crossed this part: you could be a tourist at home.’75 The second, underlined part renders the verbless Greek phrase end¯emos apod¯emia, which is impossible to translate literally: end¯emos is an adjective meaning ‘in the polity’, apod¯emia a noun meaning ‘being away from the polity’. The most obvious way of taking the phrase, followed in all translations of which I know (including my own), is as given above: Clitophon is claiming that the city is so large that those in their own polity might feel abroad. But there is a problem here. Clitophon is not at home, end¯emos, indeed quite the opposite: elsewhere, he refers to his time abroad as an apod¯emia (2.27.2, 2.33.3, 5.10.3, 8.5.7).76 The person who is at home is, of course, Achilles himself. end¯emos apod¯emia brings together, in a single, 70

71 72

73 74 75 76

sunhntto eÉqÆv tv p»lewv ˆstr†pton t¼ k†llov, kaª mou toÆv ½fqalmoÆv –g”misen ¡donv, 5.1.1. Morales comments that ‘It is a fine comic touch that Clitophon is dazzled when he walks through the gates of the Sun’ ((2004) 104). katastr†ptei mou toÆv ½fqalmoÅv, 1.4.2. o¬ m•n dŸ Šlloi ti gastrª metržsantev tŸn ¡donžn, –gÜ d• . . . tän tv k»rhv prosÛpwn gemisqe©v, 1.6.1 (alluding, as commentators all note, to Dem. 18.296). Subtler echoes: the gates of Selene (5.1.2) look back to the comparison of Leucippe to a picture of Selene (1.4.3); the ‘row of columns’ (ki»nwn Àrcatov, 5.1.4) picks up the ‘chorus of columns’ (c»rwi ki»nwn) in Hippias’ garden (1.15.1). See further Whitmarsh (2009b) 44–7. Clitophon is, for sure, Tyrian not Sidonian, but the two cities are always closely associated. ¾d¼v d• di‡ toÓ ped©ou pollŸ kaª ›ndhmov ˆpodhm©a, 5.1.3. See, however, 5.15.1, where Clitophon describes his journey from Alexandria to Ephesus as an ˆpodhm©a. O’Sullivan 1980 39 classes this as a solitary case referring to a voyage ‘from a place that is not one’s homeland’; we might, however, put this down to the subjective assimilation between author and narrator discussed above (i.e. this journey would have been a conventional ˆpodhm©a for Achilles).

Transforming romance

85

giddyingly paradoxical phrase, both Clitophon’s and Achilles’ perspectives. In this narrative, no one ever quite feels at home. The dominant figure for cultural reading in Leucippe and Clitophon is not Selden’s syllepsis, which implies the availability of two equally weighted, alternative perspectives. It is, rather, metalepsis. This term, which I take from G´erard Genette, refers to the interpenetration of different narrative levels – as, for example, in the movie Stranger than fiction, where a man persuades his author not to kill him off.77 There is nothing so flagrantly transgressive in ancient fiction: what I have in mind is a softer metalepsis, whereby the ‘sacred frontier’ that separates primary from embedded narrative becomes permeable.78 Achilles’ romance does not, as Selden suggests, identify and compartmentalise discrete cultural perspectives; rather, it conflates and confuses them. problematising narration Leucippe and Clitophon is, indeed, a text in which narratorial reportage is always difficult to assess. We learn to suspect every narrator, and therefore every stage of narrative transmission. In this respect, it has much in common with other contemporary or near-contemporary fictional texts, where narrators are subjected to scrutiny. A prime example is Lucian’s True stories, which is in form an eyewitness account of the author’s travels, but the prologue insists that it deals with ‘of things that I have neither seen nor experienced nor heard from anyone else, in fact things that do not exist at all, nor could they in the first place’.79 A tradition of Homeric revisionism emerges (rooted, to be sure, in Hellenistic writers like Hegesanax, Euhemerus and Dionysius Scytobrachion; and before that in the revisionist tradition of Stesichorus, Herodotus and Euripides) that ironically questions the veracity of the Homeric narrator: examples include Dio Chrysostom’s eleventh oration (claiming that Troy really was captured), the Journal of ‘Dictys of Crete’ (supposedly an eye-witness account of the war), and Philostratus’ Heroic tale (in which Homer is said to have been persuaded by Odysseus to whitewash him, in return for information 77 78

79

Genette (1980) 234–7, and esp. Genette (2004), which nuances the figure as la m´etalepse de l’auteur. Excellent discussion at Fludernik (2003); for classical applications see de Jong (2009). Fludernik (2003) in particular emphasises the pre-postmodern heritage of metalepsis; De Temmerman (2009b) interestingly explores a case of metalepsis in Achilles Tatius, arguing that the ecphrasis at 1.19.1–2 is interfered with by the description of the initial painting. The phrase ‘sacred frontier’ is Genette’s ((1980) 236). perª æn mžte e²don mžte ›paqon mžte par’ Šllwn –puq»mhn, ›ti d• mžte Âlwv Àntwn mžte tŸn ˆrcŸn gen”sqai dunam”nwn, VH 1.4.

86

Part i Returning romance

on the war).80 Texts like these rely on the premise that Homer is not an authoritative, muse-inspired source, but a human narrator, both partial and corruptible.81 Related is the trend towards pseudo-documentarism, the invention of fictional sources:82 in revisionist accounts, the authority of Homeric narration is undercut by documents such as the Phoenician text that supposedly underlies ‘Dictys of Crete’; or the Egyptian pillar inscription that (according to the priests Dio claims to have consulted, 11.38) gives the true story of the Trojan War; or the vintner’s consultation of the epiphanic hero Protesilaus in Philostratus’ Heroic tale. At the same time as it purports to authenticate the narrative it buttresses, however, pseudo-documentarism also ironises it:83 it is unlikely that all ancient readers really believed that Dictys represented a transcription of a discovered manuscript, any more than modern readers of The name of the rose do; or that Dio really saw an Egyptian pillar (particularly when the device so clearly recalls the literary precedents of Herodotus and Euhemerus);84 or that Philostratus’ vintner did literally encounter Protesilaus. Pseudo-documentarism does not simply validate texts, although it may do in the eyes of the naive; for other readers, it simply multiplies and relativises the sources of narrative authority, and thus militates against the possibility of any final truth. It is the product of a culture in which narrative authority is not arrogated to a single ‘master of truth’, but competed for by multiple masters of persuasion. This is particularly evident in the case of Dio 11, which emphatically shifts the question from ‘what is the truth of the Trojan War?’ to ‘what is a plausible account of it?’ (11.16, 20, 55, 59, 67, 69, 70, 92, 130, 137, 139). The request to trust 80

81 82

83 84

Dio 11 is discussed by Kindstrand (1973) 141–62; Seeck (1990); Anderson (2000) 152–3; Sa¨ıd (2000) 176–86; Gangloff (2006) 118–36. On Dictys, and the revisionist tradition in general, see Merkle (1994). For Philostratus’ Heroicus, see the essays in Aitken and Maclean (2004), esp. Mestre (2004) on Homeric revisionism; also Whitmarsh (2009a). There have been a number of recent translations and commentaries on the Heroic tale, of which the most comprehensive is Grossardt (2006), who discusses Homeric revisionism at 96–120. On the cultural context of Homeric revisionism see esp. Zeitlin (2001). I have not included Dares in this list, as I am not convinced that (despite the claim in the prologue) a Greek original lies behind the extant Latin. On the Greek Dictys, see above, n. 15. Imperial texts are distinctive for their attacks on the personal authority of Homer: see Kim (2010), a magisterial discussion. Speyer (1970) 43–124; Hansen (2003); n´ı Mheallaigh (2008). Pseudo-documentarism of this kind is, however, not new to the imperial period: earlier instances include Hegesianax’ forged account of Cepahl(i)on of Gergitha, and the inscribed gold columns alluded to by Euhemerus (Diod. Sic. 6.1.7–10). n´ı Mheallaigh (2008) 404: ‘In increasingly self-conscious fiction, such Beglaubigungstratagien are converted also into signals to the knowing reader, playfully advertising the fictionality of the text’. Herodotus hears the ‘true’ story of Helen from Egyptian priests (2.112–17); for Euhemerus’ columns see n. 82.

Transforming romance

87

Dio rather than Homer harms narrative authority in general more than it promotes Dio’s in particular. Homeric pseudo-documentarism is a rich example of what Jacques Derrida calls ‘supplementarity’.85 The supplement occupies an ambiguous role in relation to the original: it can be seen either as a necessary addition, required for completion, or as something that supplants and displaces it. Dictys, Dio 11 and Philostratus’ Heroic tale all claim to complete the original truth of the Trojan war, by reaching back earlier than the established narrative tradition to sources contemporaneous with the events themselves; but at the same time, they also represent merely the chronologically newest additions to the huge cluster of Homeric interpretations, and in that respect represent, rather than a return to the original, an accretion that obscures it further. It is in this context that we should locate the most extraordinary example of second-century narratorial problematisation, namely Antonius Diogenes’ Wonders beyond Thule (which now survives only in fragments and Photius’ ninth-century summary). According to Photius, this enormous work (twenty-four books) contained up to eight levels of embedded narration.86 It also employed a romance twist on the pseudodocumentarism, offering two non-complementary accounts of the genesis of the text. In a prefatory letter to Faustinus, the author claims (so Photius tells us) that ‘even if he invents implausible lies, he does have testimony from the ancients for most of his fictions’.87 In a second letter, however, addressed to his sister Isidora, he claims that the text was discovered, when Alexander sacked Tyre, on cypress-wood tablets that had been laid in a tomb: these represented the autobiography of one of the principle characters, Deinias. The story was then ‘transcribed’, or perhaps ‘translated’, by one Balagros.88 Antonius seems to have blended the kind of strategy found at the outset of Lucian’s True stories (admitting fantastic lies, but following literary precedent) with that of Dictys (chance discovery of an ancient text). This competing set of epistolary authentications points to an astonishing level of self-consciousness, an awareness of the text as both a physical and an intellectual construction.89 It also reinforces, once again, the central role of ‘soft’ metalepsis in second-century fiction: readers of 85 87 88 89

Derrida (1974) 269–316. 86 Discussed at Stephens and Winkler (1995) 114–16. e« kaª Špista kaª yeud pl†ttoi, ˆllì oÔn ›cei perª tän ple©stwn aÉt muqologhq”ntwn ˆrcaiot”rwn martur©av, Bibl. cod. 166 111a. metagray†menov (Bibl. cod. 166 111b), an ambiguous verb (LSJ s.v. 1–2, appositely citing Luc. Hist. conscr. 21: metagr†yai [from Latin] –v t¼ ëEllhnik»n). An awareness paralleled, what is more, within the narrative proper: see SW 149 on PSI 1177.

88

Part i Returning romance

this text will have been encouraged to read it simultaneously as a series of first-person accounts of personal experience (what Genette would call homodiegetic, intradiegetic narration) and as a work of fiction on the part of the author. If we are to think of the main narrative of the Wonders as ‘translated’ from the Phoenician, then the connection with ‘Dictys’ is closer still: the preface to the Journal claims that the work is a translation into Greek from the Phoenician undertaken in the time of Nero. The Greek text (which survives only in fragments) is thus translated once;90 whereas the Latin text, the only complete version, also features an initial letter by one ‘Septimius’ explaining that he has translated the Greek translation.91 Both the Dictys author and Antonius are presumably picking up on the idea of Phoenician as an originary language, older and more authoritative than Greek – an idea promulgated by historians like Menander of Ephesus (who claimed to have learned the language)92 and Philo of Byblos, whose claim to have translated the work of one Sanchunyaton (who lived at the time of the Trojan War) itself looks suspiciously pseudo-documentarist.93 Much as the Phoenician ur-text represents a claim to authenticity, however, it might also have connotations of scurrility and deceit: an ambience of lowlife crookedness surrounds literary Phoenicians from Homer onwards (Od. 14.288–9, 15.415–16, 419),94 a reputation still active in the imperial period,95 and no doubt mobilised in Lollianus’ now-fragmentary Phoenician affairs.96 Translation is the supplementary gesture par excellence, a return to original sources that represents at the same time a total transformation of them; and it seems that specifically Phoenician translation amplifies the supplementarity, connoting both primeval truth and deceitful sharp practice. The second century, then, sees a generally intensified questioning of narrative authority, and correspondingly an increasing emphasis upon the 90 91 92 93 94

95 96

Above, n. 3. The epistolary preface to the Latin Dares also claims that the text is translated from the Greek (by Nepos!). Menander’s Phoenician learning: FGrH 783 T 3(a)–(c); Although there is also historiographical precedent, in the not implausible claims of Ctesias, Manetho and Berossus, to have used source material in Persian, Egyptian and Akkadian respectively. On Homer’s presentation of Phoenicians see Winter (1995) esp. 247–9, noting that the negative image is counterbalanced by an appreciation of their craftsmanship. As she proceeds to argue, the portrayal of Phoenician craft and craftiness may be calqued on the image of Odysseus himself (256–8). Cf. Philostr. Her. 1.3, an allusion (diab”blhsqe) to Homer (Od. 15.416; cf. 14.289) via Plato (Rep. 436a). Morales (2004) 50, who links Leucippe and Clitophon to this tradition of Phoenician scurrility.

Transforming romance

89

limitations of individual narrators. To be clear, we are speaking here of a general shift of emphasis, rather than a sudden, sharp break. The devices noted above can (as I have made clear) all be found in texts that predate the second century; what is distinctive about our period, however, is the distillation and concentration of these techniques in specific texts like Antonius Diogenes, Lucian’s True stories and (presumably) the Greek Dictys. corrupt narration In Achilles Tatius too, the Phoenician context is significant: for reasons given above, second-century readers are likely to have approached a Phoenician story with anticipations of both sacred antiquity and deceptive depravity. More specifically, Leucippe and Clitophon is also a multiply embedded tale: there may be no pseudo-documentarism, but there is a chain of transmission, a narrative audit trail. We have already mentioned the primary narrator, whose distinctive perspective colours his ways of seeing. As an er¯otikos, he translates all that he sees into a collage of innuendo and reminiscences drawn from the Greek tradition of literary erotica. Let us briefly consider three further ways in which his perspective shapes the narrative. First, his ecphrasis of the Europa painting again emphasises the erotic aspects of the scene,97 and overlays motifs borrowed from the Greek tradition (particularly Moschus’ Europa).98 Second, the very decision to begin 97

98

Europa’s navel, belly, hips, genitalia and breasts are emphasised (1.1.11), and even the foliage is said to ‘mingle’ and ‘embrace’ (ˆnem”mikto . . . sunpton . . . sumplokž, 1.1.3). Talk of ‘meadows’, ‘horns’ and ‘foam’ is provocative, too. See Henderson (1991) 136, 127 respectively for leimÛn = female pubes and k”rav = erect penis; Šfrov = semen is common (LSJ s.v. 2, and esp. West (1966) 213); Aphrodite’s name is sometimes so etymologised (Corn. ND 45; Nonn. D. 13.439–40). The eroticisation of flora and fauna also looks forward to the garden scene beginning at 1.15 (De Temmerman (2009b)). On the eroticism of the (description of the) painting, see further Bartsch (1989) 48–51. von M¨ollendorff (2009) 157–8 rightly links the eroticised pleasure in viewing to an aesthetic dimension, noting the multiple references to the language of construction and creation (t”cnh, graf-). On the wider role of the Europa narrative, and particularly its connection to the plot, see Reeves (2007) and von M¨ollendorff (2009). Moschus’ flower catalogue (narcissi, hyacinths, violets, herpylli, roses: Eur. 65–71) is subtly transformed in Achilles’ description (narcissi, roses, myrtle: 1.1.5): Campbell (1991) 71 (too contemptuous of Achilles’ ‘half-hearted effort’). The similarities have been often asserted (e.g. Mignogna (1993) 180–1), but not systematically discussed, to my knowledge. Verbal echoes: ‘She was seated on the ox’s back’ (–pek†qhto to±v nÛtoiv toÓ bo»v, Ach. Tat. 1.1.10) ∼ ‘she sat on the ox’s back’ (–fezom”nh . . . bo”oiv –pª nÛtoiv, Eur. 125); ‘holding onto the horn with her left hand’ (ti laii toÓ k”rwv –com”nh, Ach. Tat. 1.1.10) ∼ ‘with the one hand she held onto the bull’s long horn’ (ti m•n •cen taÅrou dolic¼n k”rav, Eur. 126, where (a) the phrasing ultimately derives from Hom. Il. 23.780 ([Ajax] k”rav met‡ cersªn ›cwn bo¼v ˆgraÅloio; horn-seizing with a single hand is, however, a widespread motif in Hellenistic poetic accounts of bull-wrestling, perhaps impelled by Call. Hec. fr. 258 Pf. = 67 Hollis: cf. Ap. Rh. 3.1306–7, [Theocr.] 25.145–6, AP 16.105), and (b) dolic»n implies the ship simile that Achilles will further develop (Campbell (1991) 110); ‘the folds of her robe were stretched out, swollen at every point’ (¾ d• k»lpov toÓ p”plou p†ntoqen

90

Part i Returning romance

a tale of East–West interaction with a story of female abduction story is traditionally Greek: the obvious examples are the Trojan War (and, relatedly, the wrath of Achilles, motivated by Agamemnon’s theft of Briseis) and Herodotus’ Histories, where the Persian and Phoenician explanations for the origins of Greco-barbarian hostilities are directly linked to reciprocal thefts of women (among them Europa). 99 Finally, his description of the landscape where he meets Clitophon borrows heavily from Plato’s Phaedrus, the most clich´ed setting for erotic narrative possible.100 The initial narrator, then, certainly does seem to colour the narrative with his own kaleidoscopic reminiscences of the Greek101 erotic tradition: this encourages us to suspect that the narrative has been filtered. Then we turn to Clitophon himself, the most unreliable of narrators. How can we trust someone who starts with the claim that ‘my story resembles fiction (muthois)?’102 Clitophon is repeatedly shown up as a frustratingly flittish reporter.103 At 1.6.6, he tells how he walks around the house pretending to read a book and peeking up at Leucippe. Readers, of course, would love to know what this book is: the temptation to take this passage as a self-referential commentary upon the reading of erotic literature is almost unbearable.104 But the point is that the book is precisely what Clitophon is not interested in, and we are left frustrated.105 Similarly later in this book, the transition from the mention of the burial of Clinias’ boyfriend to Clitophon’s resumption of his pursuit of Leucippe is so abrupt

99 100

101

102

103 104

–t”tato kurtoÅmenov, Ach. Tat. 1.1.12) ∼ ‘the folds of the deep robe of Europa above her shoulders were filled’ (kolpÛqh d’ ßmoisi p”plov baqÆv EÉrwpe©hv, Eur. 129); ‘she used her robe like a sail’ (ãsper ¬st©wi täi p”plwi crwm”nh, Ach. Tat. 1.1.12) ∼ ‘[the robe is used] like the sail of a ship’ (¬st©on o³† te nh»v, Eur. 130). Hdt. 1.1–4. On the motif of woman-stealing as an aetiology for war, and the celebrated parody (?) at Ar. Ach. 528–9, see Lang (1972) and now Wright (2007) 414–7. Pl. Phaedr. 229a–30c (particularly beloved of Achilles: Trapp (1990) 155). The clich´e is already acknowledged at Plut. Am. 749a; Clitophon’s reference to the location as a t»pov ¡dÅv (1.2.3) implicitly plays upon the idea of a literary topos (LSJ s.v. 4). And possibly Latin? Like Achilles’ unnamed narrator, Vergil’s Aeneas survives a shipwreck. Coming ashore, he finds a temple for a goddess, covered with artworks, which the narrator describes as Aeneas surveys them (Verg. Aen. 1.441–93). Most suggestive of a direct link is the figure of ‘Sidonian Dido’ (as Vergil styles her: 1.446) herself. There is, indeed, something eminently Europa-like about the westward flight from Phoenicia of this descendant of Agenor (1.338), with her Astarte-like features (Hexter (1992) 348–9). It is not, in my view, impossible that Achilles read Latin (a separate study of this is needed); but in any case, Greek translations of the Aeneid did exist, albeit our papyri are late-antique (Fisher (1982) 183–9). t‡ . . . –m‡ mÅqoiv ›oike, 1.2.2. Morales (2004) 53–5 emphasises the ambiguous use of the Platonic logos-mythos distinction: ‘my story is true but resembles fiction’ or ‘my story has the properties of fiction’? Whitmarsh (2003); Morgan (2007a). On Clitophon’s ‘blindness to himself and others’, see Morgan (1997) 182–5, at 182. See e.g. Goldhill (1995) 70–1; Morales (2004) 79. 105 See further Whitmarsh (2003) 199.

Transforming romance

91

and hasty (‘After the burial, I immediately set off hurriedly to find the girl’,106 my emphasis) that one critic has suspected a textual lacuna;107 but the point is that Clitophon is once again proving himself a brutally self-interested narrator, with no concern for pederasty or the suffering of others.108 The effect of reading Leucippe and Clitophon, distinguished among the extant romances for its character-bound, homodiegetic narrative presentation,109 is like watching a secret camera film for a viewer used to Hollywood widescreen panoramas: we see only what the focaliser, Clitophon, wants us to see, and we are left more or less guessing what is happening in the penumbra. What is more, Clitophon is evidently capable of refashioning a narrative to suit his own agenda. Towards the end of the text, Leucippe’s father Sostratus asks him to recap the story for him (8.4–5). This episode is modelled on the analogous episode, towards the end of Chariton’s Callirhoe, where Chaereas is encouraged by his own father-in-law, Hermocrates, to retell the story in the Syracusan theatre (8.7.3–8.8.11). As we saw in the previous chapter, Chaereas’ narrative there is presented as an act of mastery over the trauma induced by his sufferings, and signifies his public acquisition of a mature identity. Achilles, characteristically (as we shall see presently) avoids the public context, setting his episode at an intimate symposium in the temple of Ephesian Artemis: Clitophon’s relationship of primary significance is with his family, not his civic community. Nevertheless, his act of narration is envisaged, as in Chariton, in terms of mastery of trauma and maturation. In language that clearly alludes to Chariton, Sostratus encourages Clitophon to surmount his embarassment (aid¯os) over the ‘grievous’ events that have befallen him.110 The act of narration consigns trauma to the past, and translates the suffering of experience into the pleasure of narrative: ‘a narration of events past provides more entertainment than grief for one whose sufferings are over’.111 Even more explicitly than in Chariton, this passage marks narration as a therapeutic act that separates narrated past

106 108 109

110

111

met‡ d• tŸn tafŸn eÉqÆv ›speudon –pª tŸn k»rhn, 1.15.1. 107 Pearcy (1978). Below, pp. 159–63. H¨agg (1971) 124–36; Reardon (1994); Morgan (2004d) 493–502. The other major work of imperial Greek homodiegetic narrative is the Ass, at least in its pseudo-Lucianic and Apuleian forms, and probably also in ‘Lucius of Patrae’ (the papyrus fragment P.Oxy. 4762 is narrated heterodiegetically). l”ge, t”knon Kleitofän, mhd•n a«doÅmenov. kaª g‡r e­ t© moi sumb”bhke luphr»n, m†lista m•n oÉ s»n –stin ˆll‡ tv TÅchv, Ach. Tat. 8.4.4 ∼ mhd•n a«desqiv, å t”knon, k‹n l”ghiv ti luphr»teron £ pikr»teron . . . , Char. 8.7.4. The symposium so far, we have been told, has been dominated by aid¯os (§n Âlon t¼ sump»sion a«dÛv, 8.4.1). tän ›rgwn parelq»ntwn ¡ dižghsiv t¼n oÉk”ti p†sconta yucagwge± mllon £ lupe±, 8.4.4.

92

Part i Returning romance

from narrating present.112 Yet Clitophon is no Chaereas. Whereas the latter takes the opportunity to create a story celebrating his own heroic virility, Achilles’ manipulative hero slyly refashions his account into one of sexual sobriety, censoring out his indiscretion: When I came to the part about Melite, I elevated my own role, reshaping the story into one of chaste self-control, although I told no outright lies . . . one of my actions in the plot alone I overlooked, namely the ‘respect’ (aid¯os) I subsequently paid to Melite.113

In other words he glozes over the fact that having refused sex with Melite for the entire time that he believed Leucippe to be dead, he succumbed immediately after discovering she was alive (replaying, more outrageously, Callirhoe’s strategic omission of her secret communications with Dionysius: 8.4.11). The euphemistic use of the word aid¯os, here translated ‘respect’, confirms the point that Clitophon is prone to refashioning language for his own purposes: when (in the passage discussed in the previous paragraph) Sostratus anticipated Clitophon’s aid¯os, it is unlikely that he was thinking of adultery behind his daughter’s back. Clitophon continues explicitly to refashion his story in accordance with his own agenda. Presently he tells us that ‘I elevated [Leucippe’s] role too, even more than I had done mine’,114 in an attempt to win her favour. He soon returns to the theme of his own chastity, with a breathtakingly captious formulation: ‘and as for me too, if there be such a thing as virginity in a man, I have retained it up to the present day, as far as Leucippe is concerned’.115 The double qualification means that his claim to virginity is once again not an ‘outright lie’ – but it is certainly misleading in the mouth of an adulterer and visitor to prostitutes (2.37.5). Given that it is Clitophon who (via the unnamed primary narrator) tells most of the story, these observations have repercussions for the entire erotic narrative. Clitophon’s self-censorship in the passage quoted above 112

113

114 115

In the analogous passage in Chariton, Hermocrates simply says that ‘the brilliant end overshadows all of the previous events’ (t¼ . . . t”lov lampr¼n gen»menon –piskote± to±v prot”roiv Œpasi, 8.7.4); it is a question of Chaereas’ stature in the eyes of others, not of his own psychological therapy. The idea that narrative can have a therapeutic role is traditional (Hes. Th. 98–103, with Walsh (1984) 22–4), but usually the effect is on others: I know of no Greek parallel for this idea of a talking cure. –peª d• kat‡ tŸn Mel©thn –gen»mhn, –xiron t¼ prgma –mautoÓ pr¼v swfrosÅnhn metapoiän kaª oÉd•n –yeud»mhn . . . šn m»non parka tän –mautoÓ dram†twn, tŸn met‡ taÓta pr¼v Mel©thn a«dä, 8.5.2–3. –xiron kaª t‡ aÉtv ›ti mllon £ tˆm†, 8.5.5. e­ tiv Šra –stin ˆndr¼v parqen©a, taÅthn kˆgÜ m”cri toÓ par»ntov pr¼v Leuk©pphn ›cw, 8.5.7.

Transforming romance

93

(8.5.2–3) is strikingly recursive. For a start, the scene alludes to the archnarrator himself, Odysseus, who seems to pass over his sexual relationship with Circe when recounting his adventures to Penelope (Odyssey 23.321). The language, moreover, seems self-reflexive: the word translated ‘plot’ is dramata, literally ‘dramas’ (the very word used to denote the romances themselves by Byzantine times);116 ‘reshaping’ is metapoiein, another knowingly technical term, used of illicit tampering with authoritative texts.117 How much tampering has gone on in the transmission of the narrative as a whole? Achilles thoroughly subverts the paradigm of the first-century romance, replacing its Hellenocentric naturalism with a tricky, elusive, metaleptic, decentred, self-subverting discourse. art and interpretation Achilles’ emphasis upon filtering and mediation is paralleled in an approximately contemporary text, Longus’ Daphnis and Chloe – at one level a very different kind of text, but, despite the generic blending of Alexandrian pastoral motifs and the absence of adventuring themes, still recognisably romantic in its emphasis on the love and maturation of a heterosexual couple.118 The two works have much in common (it is more than likely that one author read the other, although it is hard to be confident which is the prior):119 both begin with an unnamed painter visiting a sacred space, and subsequently seeing a painting;120 in either case, an interpreter arrives to explicate the painting, and the explication becomes the remainder of the narrative. Longus’ use of the painting motif is developed from Xenophon of Ephesus, who describes Anthia and Habrocomes as dedicating a graph¯e 116 117 118 119 120

Agapitos (1998) 128–32. E.g. in a marginal note at Hebrews 1:3 (ˆmaq”state kaª kak”, Šfev t¼n palai»n, mŸ metapo©ei), cited at Haines-Eitzen (2000) 110. The generic hybridity of Daphnis and Chloe has been widely discussed: see especially Hunter (1983) 59–83, Zeitlin (1990) 421–30 and Pattoni (2004). Alvares (2006) explores similarities and differences; see also von M¨ollendorff (2009) 153–6. As discussed above (p. 80), Achilles’ painting is not actually located in the temple precinct, at least as the text (which may be corrupt) stands. Longus’ painting has been widely discussed. Some have taken it as a Beglaubigungsapparat, i.e. a device to procure the reader’s belief in the plausibility of the tale: Imbert (1980) 210–1 (arguing for a specifically Stoic interpretation of the text’s aesthetics); Wouters (1989–90). The opposite view has also been maintained, namely that the author is pointing, Platonically, to the distance of his text from reality: Blanchard (1975) 40–1. A number of readers see an ironic tension between the fictionality of the text and the pseudo-historical Beglaubigungsapparat: cf. Perry (1967) 109–11; Hunter (1983) 38–52; MacQueen (1985) 133; (1990) 15–23; Zeitlin (1990) 434–5. See also Mittelstadt (1967), hypothesising parallels with contemporary painting; Philippides (1983). Kestner (1973) focuses the opposition of the spatial and temporal. Ecphraseis in the romances generally are discussed by Billault (1979), (1990); Zimmermann (1999); cf. also Debray-Genette (1980), Dubel (1990) and Whitmarsh (2002b) on Heliodorus.

94

Part i Returning romance

(a picture, or perhaps a written text) containing all their experiences, in the temple of Artemis (5.15.2). Similarly, Daphnis and Chloe are said at the end to decorate the cave of the Nymphs, set up images (eikonas), which we are implicitly invited to identify with those described by the narrator at the outset (4.39.2): Longus’ is another ‘self-begetting’ romance.121 That Longus’ narrative is the verbal transcription of an artwork creates different kinds of metaleptic effects. As in Chariton and Xenophon, the very process of the romance’s own creation, its ‘self-begetting’, is dramatised. At one level, the artwork is equivalent to the narrative. When the narrator expresses his desire ‘to compose in response to the composition’ (antigrapsai t¯ei graph¯ei)122 the Greek plays on the fact that the same word, graph¯e, can mean both ‘painting’ and ‘written text’. The narrator is offering to ‘compose’ (-graphein) in response to a ‘composition’ (graph¯e). But emphasis is also placed on the difference between the painting and the narrative: the key lexeme here is the prefix anti-appended to the verb graphein (‘compose’). I have translated it ‘in response to’, its most neutral rendering: it could also mean ‘in exchange for’ or even ‘in competition with’. In other words, it marks not the identity between painting and romance, but an indeterminate play between identity and difference. This phrase calls on us to weigh the likenesses and unlikenesses between the two media. It is important, then, to catch the nuances underlying the narrator’s account of how he came to write. ‘When I had seen and wondered at (thaumasanta) these and many other things, all erotic’, the narrator comments, ‘a desire seized me to compose in response to the composition.’123 The text emphasises the psychological processes, the (erotic and aesthetic) motivation that underlies this particular intermedia presentation. What is described is a two-stage process: initial wonder (thauma), followed by a desire to respond to the painting. As Froma Zeitlin notes, a comparable passage in Lucian’s On the hall distinguishes between the practice of ‘commoners’ and ‘men of culture’ in response to spectacles:124 the former merely gaze in silence, whereas the educated viewer ‘will try as much as 121 122

123

124

Kellman (1980); above, p. 62. ˆntigr†yai ti grafi, 1. pr. 3. The phrase also points to the artistic qualities of the narrative: its word-painting, compositional finesse, elegant structure and neat framing. See further Hunter (1983) 38–52; Zeitlin (1990) 430–6. Poll‡ Šlla kaª p†nta –rwtik‡ «d»nta me kaª qaum†santa p»qov ›scen ˆntigr†yai ti grafi, 1. pr. 2–3. I have adopted the conventional punctuation contra Reeve, who begins a new sentence after –rwtik‡. oÉc ¾ aÉt¼v perª t‡ qe†mata n»mov «diÛtaiv te kaª pepaideum”noiv ˆndr†sin, De dom. 2. See Zeitlin (1990) 432–3 n. 47; also, with different emphasis, Billault (1979).

Transforming romance

95

he can to linger and respond to the sight with speech’.125 A facility with language is the primary diagnostic of elite behaviour.126 Artworks involve the viewer in a power relationship, which has wider implications for one’s social standing: the disempowered are subdued into passive silence, mere ‘wonder’,127 while the empowered respond actively and emulously. A number of roughly contemporary texts dramatise the intellectual’s ability to translate art into language (e.g. Lucian’s Heracles, the Imagines of the two Philostratoi, and the Descriptions of Callistratus). The silence of the awestruck viewer, conversely, replicates the silence of the artwork itself: in their mute inertia, the two mirror each other.128 Pictures are incommunicative, aporetic; they require linguistic supplementation in order to signify. When Longus’ narrator describes the content of the painting, he lists a series of depictions, but without any awareness of them: ‘On it [the painting]: women, some giving birth and others swaddling, children being exposed, sheep nurturing, shepherds rescuing, youths courting, a landing of pirates, an enemy attack.’129 This sentence mimics the incomplete narrative cognition experienced by the viewer of an artwork: it can represent discrete episodes, but not the relationships between them. Painting operates in spatial dimensions alone; it cannot represent time (as Gotthold Lessing’s Laok¨oon oder die Grenzen der Malerei famously showed). Longus represents the achronic nature of the picture in terms of syntactical absence: the sentence lacks both the main verb that would organise the episodes into a coherent meaning, and the connecting particles that would articulate the relationships between 125 126 127

128

129

peir†setai d• Þv o³»n te kaª –ndiatr±yai kaª l»gwi ˆmeiy†sqai tŸn q”an, De dom. 2. For this theme in imperial Greek literature, see esp. Schmitz (1997) 91–8. The disempowering effect of thauma is a key theme of ps.-Long. De subl. 1.4: ‘the combination of wonder and astonishment always has power over the merely persuasive and pleasant. This is because persuasion is on the whole something we can control, whereas amazement and wonder exert invincible power and force and get the better of the reader’ (p†nth d” ge sÆn –kplžxei toÓ piqanoÓ kaª pr¼v c†rin ˆeª krate± t¼ qaum†sion, e­ge t¼ m•n piqan¼n Þv t‡ poll‡ –f ì ¡m±n, taÓta d• dunaste©an kaª b©an Šmacon prosf”ronta pant¼v –p†nw toÓ ˆkrowm”nou kaq©statai). The nil admirari theme is strong in Philostratus’ Apollonius: see Whitmarsh (2004a) 433–5. Cf. Ach. Tat. 3.15.6, where Clitophon sententiously discourses on the effects of awe (ekpl¯exis, closely related to thauma), rationalising the myth of Niobe by explaining that she became motionless in her shock ‘as though turned into stone’ (Þseª l©qov genom”nh); Luc. Imag. 1, where Lycinus claims to have been almost turned to stone by thauma on beholding Panthea (a woman he will later compare to various artworks), and predicts his interlocutor would also have become ‘more immobile than statues’ (tän ˆndri†ntwn ˆkinht»teron). In the latter case, there is also an obvious double entendre. guna±kev –p ì aÉtv t©ktousai kaª Šllai sparg†noiv kosmoÓsai, paid©a –kke©mena, po©mnia tr”fonta, poim”nev ˆnairoÅmenoi, n”oi suntiq”menoi, listwn katadromž, polem©wn –mbolž, 1 pr.2.

96

Part i Returning romance

them. Indeed, the only temporally significant words in the sentence are the present-tense participles, which indicate an iterative state rather than its duration. To convert the painting into narrative would be to supply a temporal structure, hence to give it meaning. The translation of pictorial into narrative representation, then, is as an act of intellectual control that also has consequences for the social positioning of the subject.130 This is the significance of the two-stage response of Longus’ narrator: initial wonder, a response available to anyone, followed by an assertion of intellectual and social mastery, ‘composing in response to the composition’: the narrator is vying, competitively, with his source.131 This power play is all the more important in that this is a rural setting, and ‘rustic’ (agroikos) in this period is conventionally used as the antithesis of ‘educated’:132 the literary reaction to the painting thus demarcates him from those around him. The elite status of the narrator is consolidated by the only ‘fact’ we learn about him, which is that he entered the grove ‘while hunting in Lesbos’:133 hunting is a typically moneyed pastime. In the course of the romance, we meet a number of other hunters visiting from the city, with little sympathy for or understanding of the countryside. In book 2, some ‘rich young Methymnaean men, wishing to pass the vintaging season having exotic fun’,134 put in nearby to hunt hares; their aggressive behaviour leads to war between Mitylene and Methymna, and to Chloe’s abduction. In book 4, Astylus (‘City-boy’), the son of the estate owner Dionysiophanes (also Daphnis’ father, it will emerge), is described in terms that markedly recall the Methymnaean playboys of book 2: ‘he set about hunting hares, as you would expect from a rich young man, devoted to luxury, who had come to the countryside to enjoy exotic pleasure’.135 A social gulf thus opens up between the primary narrator and the subjects of his narrative – in much the same way as the cultural gap that divides Achilles’ narrator from his Phoenician subject. Also comparable is the mediation between the local and the panhellenic. Achilles’ narrator, as we have seen, represents himself as interpreting 130 131

132 134 135

Pandiri (1985) emphasises the political angle; see further Sa¨ıd (1999) 97–107; Whitmarsh (2008) 77–9. This agonistic desire to outdo one’s own source is grounded in imperial Greece’s culture of literary emulation: compare e.g. ps-Long. De subl. 13.4 on Plato as Homer’s ‘antagonist’, with Whitmarsh (2001a) 59–61. 133 –n L”sbwi qhrän, 1 pr. 1. Whitmarsh (2001a) 100–8. n”oi Mhqumna±oi ploÅsioi diaq”sqai t¼n trught¼n –n xeniki t”ryei qelžsantev, 2.12.1. perª qžran e²ce lagwän, o³a ploÅsiov nean©skov kaª trufän ˆeª kaª ˆfigm”nov e«v t¼n ˆgr¼n e«v ˆp»lausin x”nhv ¡donv, 4.11.1. o³a might be taken as an internal reference marker alluding to the earlier n”oi . . . ploÅsioi.

Transforming romance

97

Phoenician cult for the sake of a Greek readership. Longus’ narrator is even more explicit: his work serves, he hopes, ‘both as a dedication to Eros, the Nymphs and Pan, and as a pleasurable possession for all men’.136 The Longan narrator presents himself as shuttling between the local and the general: transforming an object of local cult, physically embodied in a specific locale, into a panhellenic literary work destined for ‘all men’. A work, moreover, that now has a general applicability to all humanity: ‘it will cure the sick, console the grieving, remind one who has loved, and provide a preliminary education for one who has not; for absolutely no one has escaped Love, nor ever will . . . ’137 The phrasing here casts the text as a work of formal instruction,138 and thus marks a total generic transformation of the sacred tale. Set alongside this transformative, panhellenising impulse, however, is a desire to mimic the religious qualities of the original painting. The text is imagined not only as a ‘pleasurable possession for all men’, but also as a ‘dedication’ to the deities of the grove. This dedicatory function replicates the cultic role of the painting itself, ‘dedicated’ by Daphnis and Chloe in the grotto.139 The narrator is attempting to achieve two different, even conflicting aims: both to transform the local cultic painting into a panhellenic work of literature and to preserve the religious essence of the artwork.140 This paradoxical duality lies at the heart of Daphnis and Chloe. Scholarship on Daphnis and Chloe has tended to be bipolar, seeing it either as pious or as sophistic,141 but this is a work that manages to be both detached and sympathetic, ironic and sincere, sophisticated and religiose. Against the well-known passages that play to the reader’s sense of superiority 136

137 138

139 140

141

ˆn†qhma m•n ï Erwti kaª NÅmfaiv kaª Panª, ktma d• terpn¼n psin ˆnqrÛpoiv, 1 pr. 3. Longus’ Thucydidean allusions: Valley (1926) 101–2 has the details; see Cueva (1998) and Trzaskoma (2005) for interpretation. Á kaª nosoÓnta «†setai kaª lupoÅmenon paramuqžsetai, t¼n –rasq”nta ˆnamnžsei, t¼n oÉk –rasq”nta propaideÅseiá p†ntwv g‡r oÉdeªv ï Erwta ›fugen £ feÅxetai . . . 1 pr.3. Comparable is e.g. Galen’s assertion of the educational value of epitomes and aphoristic collections: ‘This kind of instruction is fitting for primary learning itself, and for remembering things that one has learned are necessary, and afterwards for reminding one of things one has forgotten’ (e­v te g‡r aÉtŸn tŸn prÛthn m†qhsin kaª e«v tŸn æn ›maq” tiv [Ýfelhqnai] mnžmhn kaª e«v tŸn æn –pel†qet» tiv met‡ taÓta ˆn†mnhsin ¾ toioÓtov tr»pov tv didaskal©av –pitždeiov, Comm. Ad Hipp. Aphorismoi = xvii.2 p.355.6–10 K). ˆn†qhma, 1 pr. 3 ∼ ˆn”qesan, 4.39.2. The painting, it is true, is ‘dedicated’ in a cave that is not mentioned in the proem, but this is hardly a decisive objection. Morgan (2004a) 147 notes the contrast, relating it to a wider distinction he perceives between aesthetic and religious approaches to the text; cf. Merkelbach (1988) 138 for the ‘two levels’ (zwei Ebene) on which the text can be read. As will be clear, my position is that we are asked to read on both ‘levels’ simultaneously. Pious: Rohde (1937), Chalk (1960), Merkelbach (1988); sophistic: e.g. Rohde (1914) 534–54, Anderson (1982) 41–9, Goldhill (1995) passim.

98

Part i Returning romance

by teasing the young lovers’ naivety,142 we need to set the passages where the natural world is presented as a medium for divine communication. Events are pregnant with meaning: a wolf attacks the sheep (1.11), billygoats fight (1.12.1), a cicada nestles between Chloe’s breasts (1.26). The inability of urbanites to read nature is dramatised at one point, where the Methymnaean raiding party fails to interpret the weird phenomena that ensue after their capture of Chloe:143 These phenomena were intelligible to anyone with any sense, namely that they were visions and sounds sent by Pan by way of signalling something to the sailors; but they could not interpret the reason, for no temple of Pan had been plundered144

Longus here explicitly describes a god’s intention to communicate with mortals through signs manifested (Pan is ‘signalling something’) in the natural world. Humans’ capacity to read these signs is problematic: that Pan is communicating is clear, but what he is communicating the sailors fail to ‘interpret’ (sumbalein, cognate with symbolon, ‘symbol’). The correct interpretation (that the kidnap of Chloe is the root cause) is given to the captain, Bryaxis, by Pan in a dream-epiphany (2.26.5). This is subsequently confirmed by the sight of Chloe wearing a pine wreath, which he takes as a ‘symbol’ (symbolon, 2.28.2). This episode constitutes a programmatic reminder that Daphnis and Chloe is a textual exegesis of Lesbian mystery cult, an expression of the power of the rustic gods as displayed in the natural world: interpretation should never be superficial. The Methymnaeans represent the aggressive rapacity of the urban world; their inability to construe divine meaning warns the text’s urban readers not to underestimate the forces that inhabit the countryside. We are learning, in the course of our textual pilgrimage, to become good readers, which means to read sympathetically. 142 143

144

E.g. their interpretation of Philetas’ euphemistic advice (2.9–11) and Daphnis’ fear of ‘wounding’ Chloe (3.19). The scene borrows elements from Euripides’ Bacchae (Merkelbach (1962) 209). The dolphins attacking the ship are borrowed from HhBacch. 48–50; Dalmeyda (1934) 45 n. 1 additionally notes that Longus is punning on delf±nev = lead weights for dropping on enemy ships. The narrative event – abduction of woman leads to divine vengeance – looks to the rape of Chryseis at the beginning of the Iliad (Hom. Il. 1.11–12; 93–6); there is perhaps also allusion to the weird phenomena that follow the desecration of Protesilaus’ cult-site at the conclusion of Herodotus’ Histories (Hdt. 9.116–22, esp. 116.3: –n täi ˆdÅtäi gunaixª –m©sgeto). sunet‡ m•n oÔn psin §n t‡ gin»mena to±v fronoÓsin ½rqäv, Àti –k Pan¼v §n t‡ fant†smata kaª ˆkoÅsmata mhn©ont»v ti to±v naÅtaiv, oÉk e²con d• tŸn a«t©an sumbale±n (oÉd•n g‡r ¬er¼n sesulžto Pan»v) . . . , 2.26.5. The phrasing suggests the riddling of archaic poets: cf. esp. Bacch. 3.85, Pind. Ol. 2.85, Pyth. 5.107.

Transforming romance

99

At the same time, however, such passages need to be set against Longus’ relentlessly ironical presentation of the countryside as a place of naivety, fundamentally lacking in the paideia (education/civility) that is so constitutive of elite urban living. This duality is Longus’ equivalent to Achillean metalepsis: throughout the text, readers find themselves constantly shuttling back and forth between empathetic identification with the rustics, attuned to nature as they are, and condescension towards their impossible naivety. Longan metalepsis forces us to adopt two incommensurable perspectives at once, of the knowing and the naive, ‘one who has loved’ and ‘one who has not’. In this elevation of the religious function, Longus is very different to Achilles. In Leucippe and Clitophon, the association between the cult of Sidonian Iˇstar and the erotic narrative is indirect at best: the painting provides the initial impetus for the narrative, but the two tell different stories (at the literal level at least).145 In Longus’ case, by contrast, the painting is thematically coextensive with the narrative, and also an object of cult, attracting those who come ‘as pilgrims to the Nymphs and spectators of the picture’.146 Parallels for this kind of literary description of sacred viewing are widely attested in the period (Pausanias provides numerous examples).147 Another sacred feature not found in Achilles is the mediatory role of the exegete (ex¯eg¯et¯es) whom the narrator seeks out to explicate the painting. Again, this kind of exegetical role can be paralleled, particularly in Pausanias’ Description of Greece, which is richly populated with ex¯eg¯etai and peri¯eg¯etai prepared to share their local knowledge with passers-by.148 Such figures might in some cases be little more than tourist guides,149 but in other instances they are counted among the sacred personnel of the 145 146

147

148

149

For the subtle proleptic qualities of the painting, see Bartsch (1989) 40–5; Morales (2004) 38–48; Reeves (2007). tän m•n Numfän ¬k”tai, tv d• e«k»nov qeata©, 1 praef. 1. For ¬k”thv = ‘pilgrim’ cf. LSJ s.v., and Naiden (2006) 94 (although Longus does not say that the pilgrims come to heal love-sickness: that is rather the narrator’s wish for his own work, and a reworking of Theocr. 11.1–3 (also ∼ Long. 2.7.7)). On the importance of the visual focal point in imperial religious pilgrimage see Petsalis-Diomidis (2006), and specifically on art Elsner (1996). Whether Pausanias can be counted a pilgrim remains a matter of debate (for the positive case see Elsner (1995) 125–55; Rutherford (2001); Hutton (2006) 295–6); I prefer to see the pilgrim (in the extended sense) as one of the many roles Pausanias adopts, rather than as his dominant identity. Paus. 1.13.8, 1.31.5, 1.34.4, 1.35.8, 1.41.2, 1.42.4, 2.9.8, 2.23.6, 2.31.4, 4.33.6, 5.6.6, 5.10.7, 5.15.11, 5.18.6, 5.20.4, 5.21.8, 5.21.9, 5.23.6, 7.6.5, 9.3.3, 10.10.7, 10.28.7, with Jones (2001) (valuable comments too at Winkler (1985) 234–6). As is often noted, ps.-Cebes’ Tabula is the primary literary model for an exegete explaining a picture. Plut. De Pyth. 395a-b, 396c, 397d, 400d–f, 401e, with Jacquemain (1991). Strab. 17.1.29 writes of exegetes who explain sacred matters to outsiders (–xhghtaª to±v x”noiv tän perª t‡ ¬er†).

100

Part i Returning romance

cult site, perhaps even with formal religious duties.150 It is impossible to determine precisely the status of the exegete. But the most important point is that Longus places heavy emphasis upon the process of the narrative’s transmission, as it travels from lived reality to devotional painting to cultic exegete to panhellenic narrator, and finally to reader. If the first-century romancers emphasised the characters’ own retrospective retooling of their narratives, their successors emphasise second-order mediation. Romance is (deliberately) caught in the act of its own artificing, as it is filtered through different layers of creative reception. Achilles draws attention to both his narrators’ self-interested, partial narration; Longus to the chains of multimedia transmission that lead from life to its narrative representation. What this focus on mediation insists on most forcibly is the politics of reception: at every stage, listeners and readers transform the story, in accordance with their own agendas. And, of course, we too as readers are invited to reflect on our own investment in these stories (particularly in Daphnis and Chloe, where the text’s class differentials repeatedly invite us to self-disclose as elite voyeurs of rural poverty, like the Methymnaean playboys). transforming narrative The second-century romances respond to their first-century predecessors, then, by placing much greater emphasis upon the proliferation of cultural perspectives, and upon the interpenetration of different narrative subjectivities. I want to turn now to consider closural dynamics. As we saw in the previous chapter, Chariton and Xenophon use a centre–periphery spatial model, drawn from myths of passage rites. The effect of this is to present the return as reintegration: the Greek community, compromised at the outset, is finally restored to an equilibrium that is implicitly imagined as its natural, proper state. In second-century narrative, as we have already seen, the roles of both Hellenism and the polis are greatly diminished: the experiences of the individual are central, rather than the eternal health of the city. What is more, as we shall see now, the motif of restoration, of a return to the same, is correspondingly de-emphasised; instead, the focus is upon the transformations wrought by the events of the narrative. 150

On the official status of exegetes (only attested at Olympia), see Jones (2001) 37 (making the link to Longus’ exegete); also Merkelbach (1988) 140–1. Exegetes are also attested in the actual process of cultic initiation, guiding initiands towards the true sacred meaning of their experiences (Dio Chr. 12.33).

Transforming romance

101

The most visible sign of this shift is the relocation of the marriage. In Chariton and Xenophon (and indeed in Metiochus and Parthenope, to judge from the Persian version), the lovers marry at the beginning; they are then separated, before a final reunion at the end. In the subsequent romances, with the possible exception of Iamblichus’ Babylonian affairs (although this is unclear),151 the marriage is shifted to the end. The conclusion of the romance, therefore, no longer marks the recovery of the initial state before the travels, but the transition to a new state. In line with this concern for transformation, the second-century romances develop a new language to articulate accession into erotic maturity, that of initiation. We need to be careful here, since the English word ‘initiation’ conflates what are for the Greeks two very different forms of passage rites: coming of age and entry into the brethren of a mystery cult.152 The first-century romances are centrally about the former, and not at all about the latter. In the romances of the second century and later, by contrast, er¯os is repeatedly imaged in terms of the mysteries (telos/telet¯e, mu¯esis).153 This connection itself goes back to Plato,154 but its introduction in the context of the romance marks a significant development. The marriage that (now) ends the romance is hereby marked in terms of the acquisition of a new identity. The mysteries were imagined to mark the death of the old self and the birth of the new. These effects are described by the orator Sopater (fourth century ce): ‘I saw that initiatory rite (telet¯e), which all of you initiates (memu¯emenoi) understand, and emerged from the sanctuary a stranger to myself.’155 In Burkert’s terms, ‘mystery festivals 151

152 153 154 155

Photius begins his story with the claim that they are (to use Stephens and Winkler’s translation) ‘deeply in love with each other within the bounds of matrimony, and they are in fact being married [or betrothed?]’ (n»mwi g†mou –räntev ˆllžlwn kaª dŸ kaª zeugnÅmenoi, Bibl. 74a = 190 SW). Does this mean they were actually married? There are four reasons to withhold certainty: (a) as Stephens and Winkler indicated, zeugnÅmenoi can mean ‘betrothed’ as well as ‘married’; (b) n»mwi g†mou need not mean ‘in the legitimacy of marriage’, a kind of hendiadys; it might well mean ‘in the manner of marriage’; (c) if n»mwi g†mou does refer to marriage, then it is hard to make sense of the second phrase. It is both pleonastic and bathetic to write ‘they were married, and actually they really were yoked’; (d) the next event in Photius’ summary is the attempt of the wicked king Garmos to force Sinonis to marry him, which suggests that she is not married to Rhodanes yet: although wicked kings can happily pursue married women, it would perhaps be odd to have one seeking to marry a newly-wed. Graf (2003) 4, 9, with references. Zeitlin (2008) 102–3. Such allusions are sagaciously hunted by Merkelbach (1962) and (1988); see further below on Achilles. Esp. Pl. Symp. 210a; cf. 202e-203a, 215c; fuller list and discussion at Riedwieg (1987) 2–29; see also 30–69 on the Phaedrus. xeniz»menov –pì –mautäi (‘surprised at myself’, thus Innes and Winterbottom (1988) 95, is too weak), 114.26–115.1 Walz.

102

Part i Returning romance

should be unforgettable events, casting their shadows over the whole of one’s future life, creating experiences that transform existence’.156 The Platonic analogy between sex and initiation is recurrently invoked by Achilles. Clitophon is, the unnamed narrator observes upon first meeting him, ‘recently initiated into the god’s cult (telet¯e)’.157 When Clitophon reveals his designs on Leucippe’s virginity, he asks her to allow Aphrodite to serve as their ‘mystagogue’ (2.19.1), and in the Melite episode, again, the goddess’s ‘mysteries’ are repeatedly alluded to.158 Beneath this mysteriosophic language lies an artful wordplay. The Greek word telos (cognates are used in all three contexts cited above) means both ‘ritual’ and ‘end’. In the context of erotic narratives, a ritual initiate is also someone who has experienced the end of the story. The conclusion of the narrative is indeed a telos in this double sense. Leucippe and Clitophon ‘consummated our muchprayed-for marriage rites, and went off to Tyre’,159 before proceeding to Byzantium. The verb epitele¯o (‘consummate’) suggests both the enactment of sacred liturgy and the conclusion of the narrative of romantic love.160 The first meaning is primary elsewhere in Achilles,161 but the second is also present: the adjective ‘much prayed-for’162 signals that the wedding is also the telos of the romance plot, marking the dissolution of the erotic and narrative tension. Clitophon is counterposed, in his desire for sexual initiation, to his pederastic cousin Clinias: ‘two years older than myself; he had been initiated (tetelesmenos) into the cult of Eros’;163 he is addressed as ‘an initiate for longer than me, and you are already more familiar with the mysteries (telet¯e) of Eros’.164 Clinias’ principal role in the plot is to play the ‘restraining friend’ role (compare Polycharmus in Chariton, Hippothous

156 157 158

159 160 161

162

163 164

Burkert (1987) 89. On mystery religions in general, see esp. Burkert (1987). oÉk makr‡n tv toÓ qeoÓ teletv, 1.2.2. For mystic imagery used of sex elsewhere in Achilles, see 5.15.6, 5.16.3, 5.25.6, 5.26.3, 5.26.10, 5.27.4; cf. also 8.12.4. Achilles’ initiatory motifs are discussed by Merkelbach (1962) 114–60, though his criteria for inclusion in the category are overgenerous and his interpretation overliteral. toÆv polueÅktouv –pitel”santev g†mouv ˆpedhmžsamen e«v t¼ BÅzantion, 8.19.2. For these two senses of –pitel”w, see LSJ i.1 and ii, respectively. 7.12.3: . . . Âswn oÉk –pet”lesan tŸn qus©an o¬ qewro©. Two MSS (W and M) give ˆpet”lesan, but all editors print –pet”lesan, ‘a technical term for discharging a religious duty’ (Vilborg (1962) 122). This is how I take polÅeuktov (cf. 4.17.4 (¡ polÅeuktov  Üv ˆnafa©netai); also e.g. Hdt. 1.85.8 → D.S. 9.33.2, Anth. Pal. 14.79, Xen. Cyr. 1.6.45, [Luc.] Cyn. 8, and esp. Hesych. P 2849 (t©mion, polup»qhton)). The meaning ‘accompanied by many prayers’ is theoretically possible but unparalleled. dÅo ˆnabebhkÜv ›th tv ¡lik©av tv –mv, ›rwti tetelesm”nov, 1.7.1. ˆrcai»terov mÅsthv –moÓ kaª sunhq”sterov ¢dh ti teleti toÓ qeoÓ, 1.9.7.

Transforming romance

103

in Xenophon),165 but the mysteriosophic imagery adds two components. The first is metanarrative. As an initiate, one who has already attained the telos of his own romance, Clinias is ideally placed to advise Clitophon on the ‘routes’ (1.9.7) to sexual success. It is thus to Clinias that Clitophon initially turns for an er¯otodidaskalos, a ‘teacher of desire’. Secondly, and relatedly, initiatory language introduces the play between insider knowledge and ignorance that lies at the heart of all mystery cult.166 The tension between Clinias’ knowingness and Clitophon’s ignorance is pronounced: Clitophon blurts out his sufferings, all hackneyed erotic symptoms (sleeplessness, imagining Leucippe constantly: 1.9.1–2), concluding that ‘there has never been such a misfortune’.167 This too has a metaliterary aspect: Clinias’ reply that this is ‘nonsense’ (1.9.2) bespeaks both his sexual and his literary knowingness: these are, of course, entirely standard topoi in the erotic repertoire.168 An even stronger example comes when his boyfriend Charicles announces his impending marriage. Clinias turns straightaway to literature to warn him of the risks: If you were vulgar and uncultured (idi¯ot¯es . . . mousik¯es), you would be unaware of the plots (dramata) of women. But as it is, you could even instruct others in the myths that women have supplied to the stage: Eriphyle’s necklace, Philomela’s banquet, Sthenoboea’s slander, Aerope’s theft, Procne’s murder . . . 169

Distancing himself, like many an elitist of the imperial age, from the uneducated masses,170 Clinias displays his education to make his point: in addition to the Attic tragedies checked off here, he refers elsewhere to the arch-misogynist Hesiod, to Homer and to Herodotus.171 Clinias’ status as initiate means not only that he knows about sex, but also that he knows about knowing about it. 165 166

167 168 169

170 171

Below, pp. 206–10. The Iolaus fragment (SW 368–71) draws out the pedagogical implication: Iolaus ‘learns’ (manq†nei, 3; –memaqžkei, 36) the cult secrets that are ‘taught’ (did†skein, 3; didacq”nti, 7; did†sketai, 35–6) him. oÉ g”gonen Šllo (cf. O’Sullivan (1978) 317) toioÓton ˆtÅchma, 1.9.2. For sleeplessness, see e.g. Long. 1.13.6, 1.14.4, 2.9.2, 3.4.2, 4.29.4, 4.40.3; for envisaging, Ap. Rh. Arg. 3.453–8; Virg. Aen. 4.3–5; Char. 2.4.3, 6.7.1. ˆll ì e« m•n «diÛthv §sqa mousikv,  gnoe±v ‹n t‡ tän gunaikän dr†mataá nÓn d• k‹n Šlloiv l”goiv, Âswn –n”plhsan muqän guna±kev tŸn skhnžná Ârmov ìErifÅlhv, Filomžlav ¡ tr†peza, Sqenebo©av ¡ diabolž, ìAeropv ¡ klopž, Pr»knhv ¡ sfagž, 1.8.4. For the opposition «diÛthv – pepaideum”nov see esp. Luc. Dom. 2, Lex. 24; Philostr. VA 3.43, and further Schmitz (1997) 89–91. 1.8.2 = Hes. Op. 57–8; also t¼ tän gunaikän g”nov (1.8.1) ∼ g”nov . . . gunaikän (Hes. Th. 590). Homer’s Chryseis and Briseis are referred to at 1.8.5, and Odysseus and Penolope at 1.8.6; Herodotus’ Gyges and Candaules story (important for the romancers: Tatum (1997)) at 1.8.5.

104

Part i Returning romance

In a later speech, he again connects desire with learning and sophistication: Eros, Clinias opines, is a ‘self-taught sophist’.172 What he means, on the most literal level, is that lovers can improvise on the spur of the moment. But the phrase works on another level too, by reinforcing the analogy between erotic experience and literary competence, in the form of sophistic expertise. This message is underlined by the medium: the phrase is a knowing allusion to Plato and Xenophon.173 Even the description ‘selftaught’ is ironically learned: in the first instance it alludes to Phemius, the bard who plays in Odysseus’ household,174 but it may also allude to Dio Chrysostom, who claimed – his eye glinting as brightly as Clinias’ – to be ‘self-taught in wisdom’ (a phrase itself borrowed from Xenophon).175 The phrase thus triggers a playful reflection upon the ratio of instinct and instruction involved in sexual initiation. Another twist follows shortly. When he describes sleeping with Melite in the make-shift boudoir of his prison cell, Clitophon borrows Clinias’ phrasing: ‘Eros is a self-made, improvising sophist.’176 It is a delicious paradox that this celebration of the organically creative power of desire itself rests upon the reuse of a distinctive phrase that this selfsame romance has already rendered a clich´e (and a phrase that itself has, as we have seen, a hypotextual history). Erotic initiation, then, is a game with high stakes: it implicates the reader too in the quest for both literary sophistication and social prestige. Like Achilles’ Clinias, Longus’ erotic teachers, Philetas and Lycaenion, have mystagogic aspects.177 Both claim to be divulging god-sent truths. Philetas has witnessed an epiphany of Eros: ‘I have come to reveal (m¯enuein) to you what I have seen’, he comments, ‘to announce to you what I have heard’.178 The language of ‘revelation’ suggests the mysteries, and there is also an echo of the cultic ‘things said’ (legomena), ‘things shown’ (deiknumena) and ‘things done’ (dr¯omena). His disquisition on the nature and power of Eros (2.7) is the most powerful and authoritative statement of the erotic 172 173 174 175 176 177

178

aÉtod©daktov . . . sofistžv, 1.10.1. Pl. Symp. 203d; Xen. Cyr. 6.1.14. Eros as a teacher (didaskalos) occurs in Euripides’ first Hippolytus (fr. 430 N2 ). Od. 23.347; cf. also Pind. Ol. 2.86–7. aÉtourgoª tv sof©av, Dio Chr. 1.9 ∼ Xen. Mem. 1.5; on Dio’s slipperiness here see Whitmarsh (2001a) 161. aÉtourg¼v . . . ¾ ï Erwv kaª aÉtosc”diov sofistžv, 5.27.4. See Merkelbach (1988) 164–6, 176–8, though much of his evidence is far-fetched; also MacQueen (1990) 53–4, 73–4. Chalk (1960) sees Daphnis and Chloe as a devotional work; Rohde (1937) is more measured, seeing the text as a second-order literary expression of a religious truth. See further n. 141. For the structural pairing of Philetas and Lycaenion, see Stanzel (1991) 161–2 and Morgan (2004a) 209–10 (also noting her links with Dorcon and Gnathon). ¤kw d• Ëm±n Âsa e²don mhnÅswn, Âsa ¢kousa ˆpaggelän, 2.3.2.

Transforming romance

105

principle driving the narrative, and it focuses upon cosmic theology (his first words are ‘Eros is a god’, 2.7.1).179 Although Philetas’ instruction is incomplete (focusing, perhaps, on the things said at the expense of the things shown and done), it is a nevertheless a crucial formative moment in the education of the young lovers.180 Parallel to the Philetas episode in book 2 is the Lycaenion episode in book 3:181 she provides the acts (erga) of Eros to complement the words (onomata) supplied by Philetas.182 This episode is heavily marked with the language of teaching,183 which can be used in mystic contexts, but here seems more neutral, reactivating the association between sex and knowledge that the narrator drew in the preface (1 pr. 4: the text ‘will provide a preliminary education’, propaideusei). If Philetas’ erotic instruction was coded as religious, this appears to be a more practical lesson. The only soteriology here is fraudulent: she handles her young charge ‘as though she were truly about to teach (didaskesthai) something great and truly heaven-sent’ (the narrator thereby implying, of course, that it is not).184 Between the two of them, Philetas and Lycaenion represent the ambiguity of Longan narrative, which equivocates metaleptically between the sacred and the profane. The crucial point, however, relates to the metanarrative significance of this language of education and instruction. Philetas and Lycaenion are instruments of narrative teleology, in a double sense. In terms of the kinetics of plot, they supply knowledge that Daphnis and Chloe need in order to achieve the consummation that they crave. But they also, in their states of superior knowledge, point to the lack that the lovers must make good before the plot can complete. This cognitive lack is remedied in conclusion, as ‘Daphnis did one of the things that Lycaenion had taught (epaideuse) him, and then for the first time Chloe learned (emathen) that what happened in the woods were just shepherds’ games’.185 Like all acts of closure, this final act of instruction has powerful implications for the identities of the central figures. Here, the pedagogical relationship also 179 181 182

183

184 185

Morgan (2004a) 177–84, esp. 179–80. 180 paideÅsav, 2.8.1; paideutžrion, 2.9.1. On the parallelism see Morgan (2004a) 209–10 (also noting her links with Dorcon and Gnathon). I argue at Whitmarsh (2005f) that the portrait of Philetas mobilises readers’ awareness of stories that the Hellenistic poet Philetas/Philitas of Cos was sagacious, sometimes without reward, in his pursuit of words. didaxam”nhn, 3.17.2; maqhtžn; did†xw, 3.17.3; did†xai tŸn t”cnhn, 3.18.1; did†skesqai, 3.18.2; paideÅein, 3.18.3; maqoÓsa, 3.18.4; –paideÅse, 3.18.4; –rwtikv paidagwg©av; pepa©deuto, 3.19.1; maqe±n; –pa©deuse, 3.19.2; ˆrtimaqžv, 3.20.2. ãsper ti m”ga kaª qe»pempton ˆlhqäv m”llwn did†skesqai, 3.18.2. ›dras” ti D†fniv æn aÉt¼n –pa©deuse Luka©nion, kaª t»te Cl»h präton ›maqen Âti t‡ –pª tv Ìlhv gen»mena §n poim”nwn pa©gnia, 4.40.3.

106

Part i Returning romance

naturalises social hierarchies (coinciding as it does with the consummation of marriage, imagined as the acculturated form of ‘natural’ sex). This act of teaching also marks Daphnis’ final assumption of a position of dominance over Chloe, as he becomes the teacher (just as Xenophon’s Ischomachus constructs his hierarchical relationship to his wife in terms of education).186 Daphnis and Chloe and Leucippe and Clitophon self-consciously revise the paradigms established by Chariton and Xenophon in a number of ways. There is certainly (as much criticism has emphasised) much greater store laid by ‘literary’ self-reflexivity, signalled by the use of ecphrastic description, intertextual depth, the use of philosophical and literary-critical terminology, the deployment of Atticising diction, and so forth.187 What I have sought to show in this chapter is that these features are not simply epiphenomenal responses to shifts in literary aesthetics (the emergence of the so-called ‘Second Sophistic’), but part of a larger package of narrative reorientations, all expressing a revised conception of the paradigms of identity subtending the romance form. It is a question not just of ‘literariness’, but of an awareness of the mediated, self-consciously non-natural status of narration. The focus on the text as an artefact directs our attention towards the constructedness not just of narrative itself, but also of the stories that we choose to tell about who we are. Whereas Chariton and Xenophon see identity as (fundamentally) a stable core of selfhood that can be recovered within traditional civic structures after times of strife, the second-century romancers see it as mutable, likely to be moulded by life. The second-century romances, then, are transformative. We can see this everywhere in the stories incorporated into the central narrative. Achilles’ Callisthenes, notably, starts out as a rogue rapist, but ends up changing his personality entirely: ‘everyone marvelled at his sudden transformation from a worse character to an entirely excellent one’.188 A subtler example is Longus’ Lampis, who is ‘forgiven’ for his actions, implying contrition (Long. 4.38.2). At a subtler level still, Achilles and Longus fill their romances with embedded narratives of mythical metamorphosis, which reinforce towards the thematic centrality of the transformation theme.189 This is 186 187 188

189

On progressive marginalisation of Chloe in books 3 and 4 see Winkler (1990) 114–18. From among myriad studies of these features see especially Hunter (1983) on Daphnis and Chloe; more generally, Billault (1991) and Morgan and Harrison (2008). ãste qaum†zein Œpantav t¼ a«fn©dion oÌtwv –k toÓ ce©ronov e«v t¼ p†nu crhst¼n metelq»n, Ach. Tat. 8.17.5. Laplace (2007) 705–42 discusses the transformation theme in Achilles Tatius, but overstates her position by arguing for thoroughgoing character change in Leucippe and Clitophon. Ach. Tat. 1.1.13 (Zeus), 1.5.5 (Daphne), 5.5.5 (Procne), 8.6.7–9 (Syrinx), 8.12.8 (Rhodopis); Long. 1.27 (Pitys), 2.33.3–34 (Syrinx), 3.23 (Echo). These narratives cannot be taken as straightforwardly

Transforming romance

107

a world in which character is not necessarily set for life: individuals can reshape themselves in the light of their experience, and others can reappraise their evaluations. This interest in the metamorphosis theme reflects not only literary trends (the popularity of Ass narratives, the collection of tales by Antoninus Liberalis), but also the newer conceptions of character change embedded in the works of biographers (notably Plutarch, whose account of Themistocles’ change of character Achilles may have followed).190 This transformative dynamic makes for greater self-consciousness about closure. As we have seen, the marriage is relocated to the end, where it can figure newly acquired identities. We have also considered the heavy exploitation of the ambiguities of the word telos (‘end’/‘initiation’), particularly by Achilles. It is in this context that we should locate Leucippe and Clitophon’s celebrated resistance to closure. The text concludes with the lovers marrying and sailing from Tyre to Byzantium together; but in the framing narrative that begins the text, which seems to take place soon afterwards,191 Clitophon is in Sidon, inexplicably despondent and without Leucippe. This puzzle has been widely discussed: particularly convincing are explanations that allude to Platonic precedent and the anticlosural, experimental aspects of the romance.192 As we have also seen in this chapter, the second-century romances are specifically concerned with the problematisation and relativisation of narrative: they dramatise the absence of final meaning, and the difficulty of locating a single cultural vantage on the narrative. But anticlosurality is not just a formal, literary choice: it also has implications for the identity politics of the romance. Marriage, it implies, is neither the absolute end of the story nor the natural destiny of the human subject.

190

191

192

paradigmatic, since they generally involve (a) females only and (b) sexual violence, in pointed contrast to the romance plot (Winkler (1990) 118–21; Morales (2004) 178–84); even so, the insistent return to the metamorphosis theme is striking. The notable cases of Plutarchan character change are Philip V of Macedon (in Aratus), Sulla and particularly Sertorius: see Swain (1989) 64–6. Themistocles’ character change: Ach. Tat. 8.17.1 ∼ Plut. Them. 2.7. ‘To judge by the look of you, it is not long since your initiation (telet¯es) into the god’s cult’ (1.2.2; the god in question being Eros, Aphrodite’s son); Repath (2005) 260–1. The Greek text is quoted above, n. 157. Fusillo (1997); Repath (2005), with full bibliography at 250–1 n. 3.

chapter 3

Hellenism at the edge Heliodorus

When day had just begun to smile, and the sun was beaming down onto the peaks, men armed like bandits crept over the summit of the hill that overlooks the so-called Heracleiotic mouth of the Nile, where it pours into the sea . . . 1

Practised readers of the romance – even after Achilles Tatius’ flamboyant opening – would have been bamboozled by the beginning of Heliodorus’ Charicleia and Theagenes. These first words unsettle. We begin with a striking, disorientating metaphor, which would become famous in Byzantine times. How do we read the day’s enigmatic ‘smile’?2 Is it benign, mocking, or threatening? More generally, what is the narrative context? Other romances open straightforwardly with diegetic material establishing the parameters of place, characters and sometimes period.3 For sure, Heliodorus gives us some orientating markers here, but they are notably hazy: temporality (just after sunrise – but on what day, why?), geography (the ‘so-called’ Heracleotic mouth of the Nile) and prosopography (‘men armed like bandits’ – but are they really bandits?). Matters do not become any clearer after this. The focalisation shifts to the bandits, as they attempt to decipher the scene before them: a laden ship, and the shore strewn with signs of feasting and carnage.4 The panorama is aporetic to them: after having surveyed the ship and the shore, ‘at a loss (aporountes) as to what had happened’.5 It is as if we had reached book 22 of the Odyssey without the earlier narrative to prepare us. They then spy a beautiful young woman and man, ‘a sight more aporetic (apor¯oteron) than 1

2

3 5

ëHm”rav Šrti diagelÛshv kaª ¡l©ou t‡v ˆkrwre©av kataug†zontov, Šndrev –n Âploiv lhistriko±v Àrouv ËperkÅyantev, Á dŸ katì –kbol‡v toÓ Ne©lou kaª st»ma t¼ kaloÅmenon ëHraklewtik¼n Ëperte©nei . . . , Hld. 1.1.1. Discussed at Whitmarsh (2005e), where references to Byzantine imitations can be found. The phrase is found once in earlier literature, in the Praeparatio Sophistica of Phrynichus the Arab (93–4 de Borries); a similar expression at Philo, De mutatio nominum 162. See Chapter 2 n. 37. 4 B¨uhler (1976); Winkler (1982) 95–106; Whitmarsh (2002b) 117–19. t¼ gegon¼v  ti pot” –stin ˆporoÓntev, 1.1.8.

108

Hellenism at the edge

109

the previous one’.6 Their aporia is ours too. Who are they, and how did they get there? The girl, whose name we learn is Charicleia, does explain their story presently (1.22.2–5) – but this explanation turns out to be a deception. The full story will not be revealed until half way through the text (5.32), when Calasiris, the Egyptian priest who has arranged the two young lovers’ elopement from Delphi and shepherded their journey through Egypt, discloses it. This perplexing beginning, deliberately resisting the reader’s ‘primordial need for certainty at the beginning’ of a narrative,7 is not just a hermeneutic game for the reader.8 It is also a calculated defamiliarisation of the genre, an affront to the naturalistic rules of the Greek romance. This text forces us to read the genre, and the Hellenocentric assumptions upon which it is predicated, through fresh eyes – literally, since the eyes of Egyptian bandits are our only guide at the opening of this text. This defamiliarisation of the genre is linked to the exceptional literary– historical position occupied by this text. Charicleia and Theagenes is antiquity’s longest, latest and arguably greatest romance. The date is uncertain, but the fourth century remains the likeliest candidate.9 Like the secondcentury romancer Iamblichus and Lucian, and like the male protagonist of Leucippe and Clitophon, Heliodorus was a Hellenised Syrian: as he reveals at the end of the text, ‘a Phoenician man from Emesa, of the race of Helios, the son of Theodosius, Heliodorus’.10 Emesa (modern Homs, in Syria) lay on the east bank of the Orontes, on the edge of the central Syrian steppe, around 70 km inland from the ‘properly’ Phoenician coast.11 Emesa had grown to rapid prominence in the imperial period, and apparently rapidly under the empire to one of the most important towns of the region. Its most famous export was the cult of ’LH’GBL (Elahagabal), a local Ba‘al-type god who achieved international prominence thanks to the Emesene empress Julia Domna (whose father had been a priest of the cult) and her son, the emperor Elagabalus (218–22), who introduced his worship to Rome. The first part of the name derives from the Aramaic ’elah¯a’ meaning ‘god’ (from the same root as ‘Allah’), but the Greeks identified it with the Greek Helios,

6 8 10 11

7 Said (1975) 49. q”ama . . . tän prot”rwn ˆporÛteron, 1.2.1. 9 See appendix. Winkler (1982) 96–9; more generally Hunter (1998). ˆnŸr Fo±nix ìEmishn»v, tän ˆfì ëHl©ou g”nov, Qeodos©ou pa±v ëHli»dwrov, 10.41.4. On Iamblichus, see above, p. 75. On ancient Emesa, see esp. Chad (1972); Millar (1993) 300–9. Greek sources are vague and inconsistent on the distinction between the coastal Phoenicians and the inland Arabs (‘Phoenician’ is, in any case, a catch-all term bestowed by the Greeks). Bowie (1998) notes that Heliodorus strategically dots various kinds of fo±nix – the phoenix bird, the date palm, the blood-red colour – through his text (see also Winkler (1982) 157).

110

Part i Returning romance

hence the form Heliogabalos.12 Heliodorus’ name and his descent ‘from the race of Helios’ are thus eloquent testimony to his Emesene origins, and may even point to a hereditary priesthood. In fact, a bilingual inscription from Athens gives us a clue as to what his Semitic name may have been: the ˇ S: ˇ ‘Abdshamash, Phoenician equivalent given for ‘Heliodorus’ is ‘BDSM ‘Servant of Sun’.13 The romance is set not in Phoenicia but in Africa;14 even so, it is dominated by Helios, the sun-god from the famous smile of daybreak, through the episodes (narrated in flashback) built around the oracle of Apollo (∼Helios) at Delphi, to the closing scenes in Ethiopia, the ancient land of the sun,15 and culminating with the author’s own reference to ‘the race of the sun’, discussed above. Helios/Apollo dominates proceedings throughout, along with his counterpart Artemis/Selene/Isis, to the almost total exclusion of the rest of the Greek pantheon.16 The solar emphasis of the romance reflects not only a debt to Xenophon of Ephesus (whom, as we shall come to see, Heliodorus treats as the romancer a` degr´e z´ero), but also the author’s background in west-Asian religion. It may not, indeed, be fanciful to see a figurative representation of Emesa in Meroe, the utopian Ethiopian kingdom where Heliodorus’ narrative culminates: both are centres of sun-cult standing on great rivers.17 It is possible too that the centrality of solar cult reflects changes in the religious landscape of late antiquity, but the dating remains problematic: estimates vary from the second to the fourth centuries ce.18 Most scholars prefer a third- or fourth-century date, which would locate Heliodorus in the 12

13

14 15 16

17

18

GBL probably means ‘the mountain’, suggesting an original identification with the cosmic mountain (Starcky (1975–6) 503–4 insists on ‘god consisting in the mountain’ rather than ‘god from the mountain’) rather than, or perhaps in addition to, the sun. There is, however, some controversy over this: see Altheim and Stiel (1966) 127–9, Frey (1989) 45–6, and for the doubt esp. Millar (1993) 304–5 (citing personal communication from Sebastian Brock). A Palmyrene stele records the Semitic form of the god’s name (Starcky (1975–6)). KAI 3 53 (= vol. 1 p. 13). It is also possible, however, that the Helio- element translates Aramaic ’elah¯a ì (as in Heliogabalus). The –dorus ending, meanwhile, seems to render the Semitic theophoric form ‘BD, ‘servant of’ (as in the modern Arabic name ëAbdullah). Is Heliodorus a Phoenician writing of north Africa, playfully recompensing for the Phoenician romance of the Alexandrian Achilles Tatius? On the solar traditions underlying Heliodorus’ representation of Ethiopia, see Lesky (1959), esp. 38. Bargheer (1999) is a thorough, if banal, treatment. Despite its ostensible contribution to a section on sun-cult, K¨ovendi (1966) provides only a pedestrian literary appreciation. The central role of Apollo/Artemis also invokes the precedent of Xenophon of Ephesus. Altheim (1942) draws useful parallels between Heliodorus’ theology and Emesene sun-cult. Heliodorus’ Meroe may be romanticised, but the city was real enough, the powerful capital of Kush (in modern Sudan). Heliodorus’ representation of the historical city is discussed by H¨agg (2000). See appendix.

Hellenism at the edge

111

aftermath of the Severan promotion of the solar cult of Elagabal.19 Rohde wanted to squeeze him into the narrow window of the reign of Aurelian (270–5), who revived Elagabal in the form of Sol Invictus.20 A slender majority of scholars, however, locate the text in the fourth century, some in the even briefer reign of Julian (361–3), who attempted to disestablish Christianity; his philosophical writings (particularly his hymn To Helios the king) share with Heliodorus an interest in neoplatonism and solar cult.21 But to seek a single determinant for Heliodorus’ preoccupation with the god who shared part of his own name is too reductive. Helios is inevitably an overdetermined figure; and there is certainly no call to believe that the romancer was closely allied with imperial ideology. Charicleia and Theagenes represents a new stage in the history of the romance. It is, for a start, extremely ambitious: substantially longer than any other extant romance (although Iamblichus’ second-century Babylonian affairs, now mostly lost, was almost certainly longer),22 and much more narratologically complex, in terms of flashback and embedded narration (although again Antonius Diogenes’ lost Wonders beyond Thule offers a precedent).23 Stylistically, it is much more elaborate and ornate.24 This presentational sophistication matches the thematic scale and scope of the work, which approaches the cosmic. Beginning on the Mediterranean coast of Egypt, the narrative takes us (partly under the guidance of Calasiris, an Egyptian priest) up the Nile, through Egypt to Ethiopia, home of solar cult. This journey is presented as a mystic pilgrimage for characters and readers alike. The erotic narrative shares these religious–mystic associations. Achilles Tatius, as we have seen, borrows the Platonic imagery of sex as initiation, but in the context of a narrative in which (for Clitophon at any rate) sexual urges are to be satisfied as quickly as possible. Heliodorus uses the same Platonic resources to a much more elevated effect: exceptionally for a Greek romance, both lovers remain ‘pure’ until they are married, in pointed contrast to the base adulterers who populate the text.25 Perhaps Heliodorus is 19

20 22 23 25

If the date is later in the third century, then the Ethiopians’ trouncing of the Persians at Syene in book 9 might be taken to invoke the successful Roman defence of Emesa against the invasions of ˇ ap¯ur I in ce 252, an event commemorated pseudo-prophetically in the thirteenth the Parthian king S¯ Sibylline oracle (where the Parthians are also named ‘Persians’). Rohde (1914) 493–6. 21 Esp. Bargheer (1999). With most scholars, I take the Suda’s bibl©oiv lq ì (i.e. 39) as an error (Suda s.v. ìI†mblicov 1), on the basis that Photius Bibl. cod. 94 only describes sixteen. Discussed esp. by Hefti (1950); Winkler (1982); Morgan (2004e). 24 Mazal (1958). On Platonic features, see esp. Sandy (1982). For sexual ‘purity’ (kaqar»thv), see 1.8.3, 1.25.4, 6.9.4, 8.9.12, 10.7.7, 10.8.2, 10.9.1, 10.22.3; cf. 7.8.6 for the inverse (Arsace). For the Cnemon story as a

112

Part i Returning romance

directly responding to Christian moralism,26 and certainly he is developing ideas already to be found in third-century neoplatonist theology, taking its place within a culture vigorously debating the relationship between sexual morality and spiritual health. Before she meets Theagenes, indeed, Charicleia is described as ‘reverencing virginity and holding it close to the immortals’27 – a phrase that is unimaginable, in its association of abstinence with godliness, in any of the earlier romancers. The narrative proper concludes with Charicleia and Theagenes assuming priesthoods and entering Meroe ‘so that the more mystic parts of the marriage could be consummated more joyously in the city’:28 in closure (‘consummated’, telesth¯esomen¯on, has metanarrative overtones),29 religious and erotic fulfilment fully converge, to the extent that the sacerdotal installation and the marriage seem to be fused into the same telestic process. Heliodorus promotes initiation from a subordinate metaphor for sex to an equipollent institution. mythic paradigms: centre and periphery As the brief discussion above has already shown, Charicleia and Theagenes is fusion of the generically distilled themes of the first-century romance (infatuation, separation, travel, oppression, reunion in marriage) and a more ambitious religious, cultural, even (in the non-technical sense) philosophical myth-history. Mythic paradigms are important to all of the romancers,30 but whereas the others use them primarily as resources for literary play, Heliodorus approaches them as the fundamental paradigm for life itself. Most important in structural terms is Homer’s Odyssey.31 The narrative

26

27 28 29

30 31

‘prolonged portrait of perverted, immoral, simply bad love’, see Morgan (1989a) 107–11, at 107; Morgan proceeds to offer the Arsace narrative as another example of ‘loving badly’ (112). As claimed by Ramelli (2001) 124–41; Morgan (2005) sees Heliodorus as part of an anti-Christian backlash. The well-known ancient tradition that the author of Charicleia and Theagenes was also the bishop of Tricca in Thessaly begins with Socrates’ Historia ecclesiastica (5.22 = Hld. T 1 Colonna; cf. Tt. 3, 14 Colonna), but the report is already treated with caution there (cf. l”getai), and few modern scholars would place absolute faith in it (cf. D¨orrie (1938) 275–6). –kqei†zousa m•n parqen©an kaª –ggÆv ˆqan†twn ˆpofa©nousa, 2.33.5; see further below, pp. 150–5. tän –pª täi g†mwi mustikwt”rwn kat‡ t¼ Šstu faidr»teron telesqhsom”nwn, 10.41.3. In the manuscript tradition, Greek works often conclude with the word TELOS (much as ‘THE END’ rolls up at the end of films); Heliodorus wittily exploits this convention, using a cognate form in the sense not of ‘conclusion’ but of ‘ritual practice’. Cueva (2004), offers some reflections: see now esp. Lefteratou (2010). See Keyes (1922) on the structural similarities, although his theories concerning the composition of the Aethiopica are suspect. More generally, see Feuillˆatre (1966) 105–114; Garson (1975); Fusillo (1989) 28–32; Whitmarsh (1998), (1999). The Memphitic episode in book 6 also rests heavily upon the Oedipus tragedies. Charicleia’s guiding (ceiragwgoÅmenov, 6.11.4) of Calasiris looks to Antigone and Oedipus at the beginning of Sophocles’ Oedipus at Colonus (an allusion missed by Paulsen

Hellenism at the edge

113

opens with the lovers on the Egyptian shores of the Mediterranean, just as Odysseus is on the shore of Calypso’s island Ogygia when the Odyssey begins; and, again as in the Odyssey, the explanation for this state of affairs is given in retrospective, embedded narration (cf. Hom. Od. 12.447–53).32 Calasiris, the narrator in question, is indeed a notably Odyssean figure: a stoical wanderer, wise and experienced but afflicted by sufferings, an artful narrator who is capable of deception and showmanship.33 These similarities are not accidental. On Zacynthus, Calasiris dreams to receives a dream-visitation from Odysseus himself. The hero castigates the priest for his failure to pay honours to him while passing near ‘the island of the Cephallenians’ (5.22.2);34 he then warns that ‘you will suffer the same troubles as me’,35 before passing on Penelope’s best wishes to Charicleia, since she exalts ‘chastity’ (s¯ophrosun¯e) above all (5.22.3). This is clearly an explicit signal to the reader that the travels and trials that befall Heliodorus’ lovers are to be read against the paradigm of the Odyssey. If Charicleia and Theagenes is a rewrite of the Odyssey, however, it is a rewrite from a distinctively non-Greek, self-consciously marginal perspective. At one point, Calasiris claims that Homer himself was an Egyptian: ‘Different peoples may attribute Homer’s origins to different places, my friend; and we can allow the wise man every city. But the truth is that Homer was a compatriot of mine, an Egyptian, and his hometown was Thebes, ‘Thebes of the hundred gates’, to borrow his own phrase [Ili. 9.383]. Ostensibly his father was a high priest, but in actual fact it was Hermes, whose high priest his ostensible father was: for once, when his wife was sleeping in the temple performing some

32 33 34

35

(1992)). When they arrive on the battlefield, their intervention is heavily modelled on Euripides’ Phoenissae: Paulsen (1992) 164–72 discusses the complex web of intertextual and metatheatrical motifs here; see also Fusillo (1989) 41–2. Calasiris’ and Charicleia’s intervention is modelled on Jocasta’s and Antigone’s arrival too late at Eur. Phoen. 1427–79: Fusillo (1989) 41, Paulsen (1992) 262 n. 90. (Diggle brackets all references to Antigone in the Euripidean passage.) Elmer (2008) 413–16 rightly cautions that Odyssean allusions arise primarily in the Calasiris sequence, and that the importance of Herodotus for the novel’s final third has not been fully recognised by scholarship. The Odyssey remains, however, the only text explicitly identified as a narrative hypotext for the wanderings and return of Charicleia (see main text). On the Homeric debt, see Rohde (1914) 474; Keyes (1922) 44. On the characterisation of Calasiris see esp. Sandy (1982). Carefully chosen words: in Homer, ‘Cephallenians’ is the generic name for different islanders under Odysseus’ command (Il. 2.631); the precise Homeric location of the specific island of Cephallenia, however, was the subject of debate (Str. 10.2.10, 13–14). I take the entire scene of Odysseus’ appearance as a Heliodorean–Odyssean refashioning of Philostr. VA 4.16, where Achilles appears to Apollonius at Troy, asking him to put right the Thessalians’ neglect of his shrine; cf. also Her. 53.19–23, with Follet (2004) 227 on the historical event behind Achilles’ ‘retribution’ (i.e. punishment for flouting imperial restrictions on the production of purple). tän ¾mo©wn –moª paqän a«sqžshi, 5.22.3.

114

Part i Returning romance

ancestral ritual, the god coupled with her and fathered Homer, who bore on his person a token of this unequal union; for, from the moment of his birth, one of his thighs was covered with a shaggy growth of hair. Hence, as he wandered his way around the world, particularly through Greece, performing his poetry, he was given the name (ho m¯eros = ‘the thigh’). He himself never spoke his true name, nor did he name his city or his descent, but the name was fabricated by those who knew of his deformity.’ ‘What was his purpose in concealing the land of his birth, Father?’ ‘It may be that he felt ashamed of being an exile, for he was banished by his father, after the mark he bore on his body had led to the recognition of his illegitimacy at the time when he came of age and was being enrolled as a priest. Or possibly this may be another example of his wisdom, and by concealing his true place of origin, he was claiming the whole world as his own.’ (3.14.2–4)

Homer’s homeland was the subject of famously intense (and wilful) speculation throughout antiquity; it was well known that a number of different cities laid claim to him.36 Enlisting the aid of an outrageous etymology,37 Calasiris now appropriates him for Egypt. Although this is neither the first nor the last appearance of this claim in Greek literature,38 its importance is double-weighted in the context of a narrative that works as a Nilotic refiguration of the Odyssey. Indeed, Homer’s biography, as presented here, bears a remarkable similarity to Charicleia’s story:39 a tale of contested paternity, foreign travel necessitated by a physical defect, and wandering.40 Homer’s life also resembles that of Calasiris, who was driven by fear of sexual scandal from his priestly Egyptian home (2.24.5–25.6). The exile theme (which will be further discussed in Chapter 6) also has a wider relevance: all of Heliodorus’ major characters suffer some kind of expatriation.41 Charicleia and Theagenes, then, can be read as the Odyssey that Homer would have written had he lived his days on the fertile banks of the Nile. 36

37 39

40 41

E.g. Dio Chr. 47.5, 55.7 (discussion at Kindstrand (1973) 113–14); Paus. 10.24.3; Philostr. Her. 18.1–3; Anth. Plan. 292–302. The Homereion of Ptolemy IV in Alexandria paraded statues representing all the poleis that laid claim to Homer’s birthplace (Ael. VH 13.22). In a gesture similar to that of Calasiris, the Syrian Lucian claims that Homer was Babylonian (VH 2.20). Possibly derived ultimately from Eur. Bacch. 294–7: Anderson (1979). 38 Sinko (1906). As noted by Winkler (1982) 102–3; Anderson (1982) 38; Fusillo (1988) 21–2; Bartsch (1989) 145. Calasiris’ account also suggests the birth of Alexander as presented in the Alexander Romance (1.4–12). Characters refer to their lives of wandering (Šlh / pl†nh): 2.24.5, 5.16.2, 5.2.7, 6.15.4, 7.8.2, 7.13.2, 7.14.7. This theme will be further discussed in chapter 6 below. Cnemon (1.14.1), Theagenes (2.4.1), Cnemon’s father Aristippus (2.9.3), Calasiris (2.25.4, 3.16.5), Charicles (2.29.5), Sisimithres (2.32.2), and Charicleia (4.18.2) are all described as exiles. Comito (1975) treats this theme, albeit without insight. See further below, pp. 220–3.

Hellenism at the edge

115

Homer, Calasiris and (implicitly) Heliodorus himself are linked by a thematic chain of exoticism, priesthood and deracination. This has significance for the geographical structuring of the text as a whole. The Odyssey is the paradigm of the centre–periphery structure that we have associated with the ephebic romance. Heliodorus’ romance, by way of contrast, is shaped around a linear north–south axis.42 The protagonists frequently express their feelings of disorientation at their seemingly random travels,43 but in fact their trajectory is unwavering and consistent: up the Nile we proceed, edging inexorably towards the exotic kingdom of Ethiopian Meroe, at the edge of the world as it was known to the Greeks. Since Homeric times, the Ethiopians had been the ‘most distant of (eskhatoi) men’ (Hom. Od. 1.23).44 Homer’s centre–periphery construction of geography is regularly echoed by characters in the romance: the necromantic corpse, for example, prophesies that Charicleia will end up ‘at the most distant (eskhatois) limits of the earth’ (6.15.4).45 Heliodorus also refers, in the reflexive self-revelation that we considered above, to the end of his narrative as a peras, a word that regularly denotes the furthest boundary of the world: ‘such was the peras of the story of the Ethiopian affair of Theagenes and Charicleia . . . ’.46 In reaching the end of this romance, the reader has also attained the limit of human ken. In one sense, then, the linear shape of Charicleia and Theagenes reverses the spatial paradigm of the Odyssean return narrative, also used by the first-century romancers, which distinguishes the geographical hub (Greek space, narrated at the beginning and end) from the non-Greek periphery (‘abroad’, narrated in the middle). In another sense, however, Charicleia and Theagenes remains a centre–periphery text: from the Ethiopians’ vantage, this is precisely a story about expatriation into an unfamiliar foreign space, and subsequent home-coming. Charicleia is the girl, says her father Hydaspes, ‘whom [the gods] exiled from her home land to the ultimate limits (perata . . . eskhata) of the earth’ (10.16.6).47 This reorientation is all the more striking when we consider that it is Delphi – for Greeks, the ‘navel’ (omphalos) or ‘hearth’ of the world – to which Hydaspes is referring. 42 43 45

46 47

Szepessy (1957) 244–54; L´etoublon (1993) 108–9; Fusillo (1989) 29. Above, n. 40. 44 See further Romm (1994) 49–54. gv –pì –sc†toiv Âroiv, perhaps a tragic quotation (it fits an iambic trimeter: Rohde (1914) 480 n. 3). Even Charicleia submits, wondering how Thisbe might have travelled –k m”shv tv ëEll†dov –pì –sc†toiv gv A«gÅptou (2.8.3); eschatic language also at 2.28.2, 4.14.2. toi»nde p”rav ›sce t¼ sÅntagma tän perª Qeag”nhn kaª Car©kleian A«qiopikän . . . , 10.41.4. . . . ¥n –xÛikisan [sc. o¬ qeoª] tv –negkoÅshv –pª p”rata gv ›scata, 10.16.6. This technique is perhaps borrowed from Ap. Rh. 3.678–80, where the Colchian Chalciope uses the language of the edge of the earth (–pi ga©hv / pe©rasi) in such a way as to imply that Colchis is central.

116

Part i Returning romance

Charicleia and Theagenes is a linear narrative for its male, Greek protagonist, but a centre–periphery narrative for his female, Ethiopian counterpart. Like the second-century romancers, but even more radically, Heliodorus plays off against each other distinct cultural perspectives, and hence forces his readers to interrogate their own. When we reach the end, we have learned to reassess our expectations, of not only the generically conventional structure of the romance, but also the received (Greek) mapping of the world. The other major mythic narrative can be dealt with more briefly. The story of Perseus’ rescue of Andromeda from an Ethiopian sea-monster is the subject of the painting that causes Charicleia to be born white (4.8.3–5; 10.14.7; see also 10.6.3).48 Given the loss of Euripides’ celebrated play (and the difficulty of interpreting Aristophanes’ lurid parody in the Thesmophoriazusae), it is hard to be certain how much direct intertextual engagement Heliodorus is undertaking. The broad themes, however, are clear from Apollodorus’ summary of the Perseus narrative in Library (2.34–49). This version is clearly ephebic:49 it narrates Perseus’ transition from early manhood, at a time of personal threat, to a recognised, propertied, settled, married state. As with Jason, Heracles and Bellerophon, when Perseus comes to manhood (¯endr¯omenou, 2.36) he is sent off on a challenging mission by a jealous ruler who wants him dead (Polydectes, the brother of the king of Seriphus, who has designs on his mother Danae). After his Gorgonslaying adventures and rescue of Andromeda, he returns to Seriphus, kills Polydectes, and ultimately founds a dynasty at Tiryns. It is also a story about the psychosexual stabilisation of gender relations, about the transition from a period of anxious defence of his (perpetually harassed) mother to the acquisition of a bride – via the destruction of a deadly female figure, who represents the terrors of female sexuality (hence Freud’s famous analysis in Das Medusenhaupt (1922), which reads Medusa as a figure for castration anxiety). The myth also manifests a clear centre– periphery structure, with the Gorgons and the sea-monster figuring the threats to be found in peripheral space. Heliodorus’ Theagenes, like Perseus, achieves manhood (as we shall see in the following section, Charicleia and 48

49

On Heliodorus’ treatment of the myth, see Billault (1981). The myth is also found in pictorial form at Ach. Tat. 3.6.3–7.9: see Morales (2004) 174–9, with references. The best-known version in antiquity was Euripides’ Andromeda, now largely lost, which featured a celebrated donna abandonata aria and an Ethiopian escape plot: for the Ethiopian setting of Euripides’ play, see Wright (2005) 129, on the basis of Ar. Thesm. 1098. For other Ethiopian versions, see Eratosth. Cat. 15–17, Apollod. 2.4.3–5, Philostr. Imag. 1.29.3. An alternative tradition, apparently later, locates the rescue in Joppa, modern Jaffa: Paus. 4.35.9, Joseph BJ 3.420, Str. 16.2.28, Conon FGrH 26 F1.40. Cf. John Tzetzes ad Lycophr. 838, 15 (pr¼v ¤bhn –laÅnontov).

Hellenism at the edge

117

Theagenes has significant ephebic elements) and rescues his beloved, before taking her home. Once again, however, the geographical model of the canonical myth has been reversed, as Ethiopia becomes the homeland and Greece the foreign space. reforming narrative The primary narrative template that Charicleia and Theagenes rewrites, however, is that of the romance – or, more specifically, that of Xenophon of Ephesus, which serves throughout as a foil for Heliodorus’ extravagant experiments. Like Xenophon’s Habrocomes, the young Charicleia holds love and marriage in contempt (2.33.4, 3.17.5).50 The account of the Thessalian ritual of propitiation at Delphi, where Charicleia and Theagenes fall in love, closely reworks Xenophon’s description of the Ephesian procession at the outset of Anthia and Habrocomes,51 also taking the opportunity to lard the scene with additional opulent, ecphrastic detail.52 This description is presented by Calasiris to the eager narrator Cnemon, whose interruptions, refusing to allow his narrator to skip over detail and controlling the rhythm of the narrative (especially 2.32.3, 3.1.1, 3.2.3, 3.12, 4.3.4, 4.4.2), have been much discussed: Cnemon is playfully set up as kind of (anti-)model reader, revelling indulgently in the sumptuous technicolour description.53 What has not been noted, however, is that Cnemon’s interventions also have an intertextual function: it is the more lavish account that they prompt that allows Heliodorus to mark the distance between his festival scene and its skeletal model. Heliodorus’ romance, twice the length of Xenophon’s, is also (its author hints) twice the romance. 50

51

52 53

Xen. Eph. 1.1.4–6; cf. also line 60 of the major papyrus fragment of Metiochus and Parthenope = SW 86 = HU 25 (with 29 n. 21). The ‘marriage rejection’ motif is also found at Joseph and Aseneth 2.1. Noted briefly by G¨artner (1967) 2080–1 (non vidi Schnepf (1887)). In Xenophon, Anthia appears first and is acclaimed by the dazzled spectators for her beauty (–kplžxewv . . . kalž, 1.2.7), before being eclipsed by Habrocomes (p†ntev «d»ntev ìAbrok»mhn –ke©nwn –pel†qonto, 1.2.8); in Heliodorus it is Theagenes who dazzles the spectators with his beauty (–xeplžtte . . . k†llouv . . . k†llov, 3.3.8), and is then ‘defeated’ (cf. ¡tthqnai, 3.4.1) by Charicleia. Xenophon’s maidens and ephebes process ‘in line’ (kat‡ st©con), carrying t‡ ¬er‡ kaª didev kaª kan kaª qumi†mata (1.2.4); in Heliodorus, the Thessalian girls preceding the ephebic procession hold a dance ‘in line’ (sticžrh, 3.2.2), bearing kan pemm†twn te kaª qumiam†twn (3.2.1); the torch (lamp†dion, d†idwn, 3.4.6) is held by Charicleia. Anthia wears a citÜn ‰lourgžv (1.2.6), Charicleia a citäna . . . ‰lourgon (3.4.2); Anthia’s girdle (zwst¼v e«v g»nu, 1.2.6) is elaborated into a lengthy ecphrastic description in Heliodorus (3.4.2–4); Anthia has k»mh xanqž, ¡ pollŸ kaqeim”nh, ½l©gh peplegm”nh, 1.2.6, Charicleia k»mh . . . oÎte p†nthi di†plokov oÎte ˆsÅndetov, 3.4.8; both have bright eyes, although there are no linguistic parallels (Xen. Eph. 1.2.6 ∼ 3.4.6). Hardie (1998); Whitmarsh (2002b) 119–21. Winkler (1982) 140–4; Morgan (1991), 95–100; Hardie (1998); Hunter (1998) 53–6; Morgan (2004e) 535–8. See further below, pp. 172–5.

118

Part i Returning romance

Again as in the first-century ephebic romances, the effect upon Charicleia and Theagenes is presented as a form of sickness (3.7.1, 3.11.1, 3.18.2, 4.5.2). Again we meet the paradox that the festival, the ritual designed to engineer social cohesion,54 leads to erotic malady, and ultimately to chaos within the community, namely when the Aenianians kidnap Charicleia. The importance of this event to the entire polity, indeed, is stressed: an assembly is called (4.19), and the general Hegesias passes motions to pursue the Thessalians and kill them, banning any of their descendants from future rites (4.20.2–3). The charge to recapture them involves all the Delphians, including women, children and the aged: ‘the entire city (polis) was keenly aggrieved at Charicleia’s abduction’,55 applying itself ‘collectively’ (pand¯emon) to the pursuit (4.21.3). This emphasis upon pandemic cohesion in the face of erotic crisis will be familiar from Chapter 1. Charicleia and Theagenes is not, however, an ephebic romance, at least in the first-century sense. The wounds inflicted upon the Delphic community are not healed at the end; Charicleia is not reaggregated into the community. In fact, Charicles does reappear at the end to demand the restitution of his ‘daughter’ (10.34.3–4); but he admits defeat, and ends up reconciled to her new (and indeed old) identity as an Ethiopian princess. This episode serves to bring into focus the whole question of true and false beginnings – and hence the metageneric question, of whether this will be a Xenophontic centre–periphery romance, or a new kind of work that images the act of initiation as a journey to a wholly unfamiliar space (a journey that is, in Charicleia’s case at least, also a kind of return). If Charicles represents the expectations of the first-century romance, the claims he stakes are partial and deceptive: ‘When Hydaspes bade him state more clearly what he meant, the old man (who was really Charicles) concealed the truth of Charicleia’s origin (genous) . . . he set out the story, summarising only the harmless parts’.56 Charicles, who embodies the homing instinct of the conventional romance, is forced into compression and falsification of the true story. The ending of Charicleia and Theagenes thus rests upon a settlement between two different generic forces, one established and the other wholly new, each represented by a different claimant to Charicleia’s paternity. The tension between the two is expressed by Theagenes, who, 54 55 56

Heliodorus’ festival is an –nagism»v, a ritual expiation of the pollution entailed by the murder of Neoptolemus at Delphi (2.34.7; see Hilton (1998) ad 3.1.1). psa ¡ p»liv Ëperžlghse tŸn Carikle©av ˆfa©resin, 4.21.3. kaª toÓ